As one example of a paper on this topic.
Probing the perils of dichotomous binning: how categorizing female dogs as spayed or intact can misinform our assumptions about the lifelong health consequences of ovariohysterectomy (freely available at pubmed)
All previous studies examining gonadal influences on canine longevity relied upon categorizing females as "intact" or "spayed" based on gonadal status at the time of death. Our study of Rottweilers generated a novel result: Keeping ovaries longer was associated with living longer. This result challenged previous assumptions that spayed females live longer.
In short - if you don't take into account that females may be spayed partway through their lives, and assume they were spayed at or before reproductive maturity, your categories should properly be named 'unspayed' and 'spayed at some point throughout their life'.
If you then go on to analyse the bitches by age of spaying, early spaying is not better.
In this study population, there was a three-fold increased likelihood of exceptional longevity (living ≥ 13 yr) associated with the longest duration of ovary exposure. However, categorizing females in this population as spayed or intact yielded the spurious, contradictory assertion that spayed females (presumed to have the least ovary exposure) are more likely to reach exceptional longevity than those that are intact.
Another paper - also openly available:
Breed Variations in the Incidence of Pyometra and Mammary Tumours in Swedish Dogs
This paper analyses the results of 260000 bitches in sweden, and the results.
In general
he proportion of the bitches that developed these diseases at 10 years of age was for pyometra 19%, MTs(Mammary tumor) 13%, and either or both pyometra or MTs 30%.
.
However, there is a WIDE variation between breeds.
There were 20 breeds with a proportion of at least 50% affected by either or both of the two diseases at 10 years of age
Looking in the supplemental data table
Supplementary data table, you can see this broken down.
For example, for pyometria, the top three risky breeds are Bernese mountain dog (66%), Great Dane (62%), Rott (61%).
Goldens are at 36%, GSD 31, 'mixed breed' 24%, Husky 14% with greyhound trailing the pack at 11%.
For Mammary tumor, dobermen and irish wolfhound start out the list at 46%, with greyhound at 7% and Husky at 5. (Mixed breed 16%).
The combined risks start out at an eyewatering 73% for Leonbergers, with huskies and greyhounds down at 16%.
So, the benefits of earlier spaying (or ovarectomy) on a greyhound may be nonexistant compared to the health costs, but this is probably not the case for Leonbergers or Wolfhounds.
Noting of course that this is only part of the picture - greyhounds are about twice as likely to need to be put down for breaking something than the above. (Mortality in North American Dogs from 1984 to 2004: An Investigation into Age‐, Size‐, and Breed‐Related Causes of Death)