Of course we all have different tastes
Yes, it's not going to be possible to create the perfect toy because of that, but I think there are ways to make ones significantly better than what is currently available. What I'm going to describe is what you might call "all in" or "premium", I don't think it would be cheap, even with regular production probably between $500-$1000 per. There is reasons for it all, and once you put a lot of effort into one feature it doesn't really make sense to stop short.
First off on the issue of firmness and geometry which constitutes a major part of the customization factor you said:
Using inflatable baloons seems impractical as it would require more pressure and more control on where the pressure is focused / distributed along the whole lenght of the toy, which would need a lot of syringes and a lot of air compartments, but i might be wrong as i am making assumptions right now.
You are exactly right it would take a lot of compartments. If you look at real anatomy there is a very fine pressurized network to accomplish engorging based erections (male and female). There is also connective tissue providing tension. In order to get that 'real feel' you need to reproduce that, not to the same degree, but to some degree.
By placing stiffer elsaters to serve as connective tissue (a stronger rubber band like fiber, but you wouldn't use rubber) in the silicone in the right pattern and by having complex internal pressure loops you could have one toy that could have a significant range of lengths and girths.
You would have independent loops for length, girth at various points, and when trying to reproduce features like a stallion's flare or a dog's knot you would use loops for them.
In order to simulate bone-like structures I would use a soft thermoplastic (not silicone, it never gets strong enough) with fibers sticking out of it to form a reliable connection to silicone and other fibers. Ban anything that would ever break into something sharp or abrasive.
I think syringes would never be an effective or reliable way to manage these pressures. I would have a computer control beyond the base with a single pressure pump and electronic pressure regulators for the various loops. I call them loops because the fluid would be in a closed circuit, it can be pressurized and then to deflate or as a safety valve be let back into a reservoir.
There are a couple reasons to use a computer:
#1 Convenience, a toy that takes an hour of prep or cleanup will rarely be used, a toy that requires a user manual to be open while you're trying to surf porn or find some sweet spot will be left on the secret shelf. It needs to be simple settings that can be remembered and activated with a press of a button.
#2 A computer can run programs relating variables, something that could be very interesting as you'll see with (see feature "Smart Toy") below.
There are also good reasons to use a closed loop:
#1 the very fine passages and crannies in the toy that produce the real feel and finely controlled geometry are not something you want to clog, you'll never clean it again. The fluid should therefore be very clean, and as many of you probably know from working on engines or computer cooling loops any time you keep changing a fluid you need a whole array of filters, catch pans, traps, etc... It's better to have one fill up (even if the consumer does need to access the fluid it should not be an everytime affair) with very clean fluid, and then reuse it.
#2 Having a loop instead of a dead end allows for some very basic cleaning action.
#3 Having a loop allows you to circulate fluid without loss of pressure (see feature "Temperature Control")
I did create a dog dildo with some internal passages (not nearly as many as I am advocating for) and it was a dead-end, It was great at first but the silicone delaminated from the fill tube and eventually something or other clogged it up. It's now permanently flaccid and unknotable.
The fluid could be gas or liquid but I would go with liquid, specifically distilled water. Why?
#1 It carries heat much better than air (see feature "Temperature Control")
#2 The resulting density of the toy would be more realistic
#3 Air is compressible, you want to be able to control the compressibility with an air column or spring beyond the base; you don't want to be locked into air-compressibility or water-incompressibility.
#4 If some kind of catastrophic failure does occur air will rise to the top and fluid will fall to the bottom, and as noted air is compressible which means it can store energy when under pressure. I personally don't think a little pressurized air escaping from a dildo would represent a serious health risk but better safe than sorry.
So I've described the basics of a pretty flexible geometry, but I would include more features because after throwing in computers, pumps, and tiny expansion tanks why not make full use of them? So far:
Core Features:
Variable length from realistic flaccid to +-20% of fully erect
Variable girth (possibly at different points) from realistic flaccid to +-20% of fully erect
Variable firmness depending on how much water vs air is in your expansion columns
Functional knots and flares
Additional Standard Features:
Temperature Control: As simple has having a heat exchanger in the clean-water pre-reservoir and slowly circulated it while the computer checks the temp going in and out. Some people may not care about this, others might care about it more; keep in mind that since water is being used the thermal mass of the erect dildo/artificial vagina would be significant unlike a pure silicone toy. Also note that some animals (like horses) are much more sensitive about it than humans. You would want option to be standard if you create models they might utilize. Finally silicone and almost all other materials involved would be able to withstand 75C water, a 'sterilize' (I know not true sterilize but pretty close) cycle could be used and would be especially useful for dildos with a hard to clean exterior surface (somebody is going to want one) or artificial vaginas.
Urethra, Ejaculation, Autolube, and Scrotum: A lot of people want functional dildos, I did; and I think it should be standard on such an (already) expensive toy. Casting in a urethra is fairly straightforward, but even here I would not simply attach the other end to a syringe.
Instead I would include multiple computer controlled "syringes" which are integrated into the clean-water pre-reservoir (not fluid contact, just to make them the same temp). I think it would be possible to place the pre-reservoir, heat exchanger, "syringes" all within a silicone scrotum.
The "syringes" would not be cylinders with a moving part. Instead they would be bifurcated cavities cast into the silicone that could be precharged with liquid (by a real syringe). Then, by shunting pressured clean water into the neighboring cavity (like any other loop), ejaculation could be accomplished. (Call these cells)
There would be two large cells for the two balls. These could be filled with the same fluid (representing semen) or two components which are meant to combine to be semen such as a cloudy one and a clear one. There would also be a cell for lubricant, this would be fixed near the top or something so it doesn't appear to be another testes.
The cells/testes could be suspended by nylon cords and pulled back with a positive pressure water hydraulic or small electric motor (but I would try to make it hydraulic the fewer systems the better) to simulate varying looseness of tendons or as part of an ejaculation sequence.
Because these cells/testes would be filled with water-like liquids and be suspended from a cord they would act pendulum like which could be desired.
Cleaning the cavities would be required, demanding non-organic simulated semen or pre-seminal fluid is unrealistic. To clean I would use a funnel which feeds into all of the cavities and run a clean program that flushes through several times. Intermittent cleaning with high saline is probably fine in most cases. To deep-clean just use boiling water (like from a electric teapot).
Smart Toy:
You can probably see a pattern here that all the important stuff is controlled by the computer. This isn't just so you don't have to reach back there and fiddle with syringes, knobs, and what not; although that is a sufficient justification for it.... that can force you into some uncomfortable positions or limit the positions you would like to have.
The computer could allow not only for remote control, presets, timed sequences, but also even more interesting possibilities. If pressure sensors are placed throughout the toys (receptive and penetrative) in appropriate places the computer could simulate responses to stimuli.
That is you could have a program that starts from a flaccid dildo or dry artificial vagina, stimulate it, have the computer observe the patterns and pressures and respond by erection, releasing lube (as if its preseminal fluid for dildos, or as if its arousal fluid for artificial vaginas), flaring, knotting, ejaculation, balls pulling up, vagina squeezing, winking, vibrating or whatever other unique features exist on the toy.
Finally the computer could be networked and the toy could be used in a multiplayer game of some kind.
Standard Mount:
All primary structure elements such as pseudo-bones, ligaments, tendons (like the nylon cords to the balls) would be routed to something stable that remains outside the body at all times, perhaps made of aluminum. This 'frame' would have a standard pattern of mounting so that if someone wanted to make a handle, fixed base, auto-thruster, inflatable, etc.. etc... the dildo or artificial vagina could be mounted firmly without complication.
In the case of an auto-thruster connecting that to the computer for simulation would be a good idea.
Cons
:
There are some downsides, these toys (especially around the base of the dildo) would be larger and heavier than anything I've seen on the market. The idea of trying to thrust into yourself by reaching one hand back probably won't work well. Women could probably manipulate it easier but still a fixed base or some kind of harnessed handle that you can manipulate with your hands would be required, especially if you buy some semi-realistically sized equine thing.
It would need power, for computer control and the pump a lithium battery would probably work, for heating probably plug in (and then maintenance with battery).
It would be designed to last but at the same time almost impossible to repair. Bigger means harder to hide, which matters for some people.
Conclusion:
There are three basic reasons people buy sex toys around here, first is to give them direct pleasure, second is to give them fantasy satisfaction, third is to prepare them for the real thing.
Those goals can overlap but they need not, the most pleasurable toy for male anal penetration is not going to be the most pleasurable toy for vaginal penetration. The most anatomically accurate toy may not be the ideal toy for pleasure at all but it does prepare you better. A swinging ball sack with liquid that will be pumped into you doesn't provide direct pleasure or practice, but it may help with the immersion of fantasy satisfaction.
