To the OP,
Not an easy question to answer. We know from the previous discussion that young athletes have historically been roughly 4 times as likely to suffer SCD as non-athletes. According to a large Italian study, screening for risk factors at a young age reduces the probability by 80%, putting athletes roughly back on par with the spectators.
Caruut produced some numbers demonstrating the death rate in pro soccer over the last 10 years remains similar to per-screening death rates, but it seems that full screening only came into force around 2006.
So, the question is, are the preventable athlete SCDs attributable to insufficient screening, PEDs, or some other factor? I suspect lack of screening has a significant role. I'm assuming the Italian athletes who were cleared for competition by the study's screening used similar amounts of drugs as their overseas counterparts. It seems to be the screening not the drugs that makes the difference.
On the other hand, there don't seem to be good data available about the current SCD rate in elite athletes, but there is an impression that there are rather a lot of these tragegdies. So, we can't rule out a recent, unmeasured and unexplained increase in fatalities.
While it is known that steroid abuse increases the incidence of SCD, there does not appear to be a known link with EPO; fatalities due to blood with the viscosity of bitumen appear to be considered as a separate cause of death. Most SCDs occur during intense exercise, so it is actually plausible that increased oxygen carriers reduces the stress on the heart and therefore reduces the risk. That is pretty wild speculation though. The simple fact that most deaths occur during intense exercise does not preclude the possibility that athletes are at increased risk due to the combination of long terms exercise induced changes in their heart structure, plus short term exercise stress.
A number of comments up thread relate to exercise induced heart changes. Hypotrophic cardiomyopathy is a thickening in one specific place, not a general enlargement. I've tried to find whether exercise can worsen any pre-existing HCM, but I'm not sure that the medical types have actually decided that. There does seem to be some indication that exercise induced changes make it more difficult to detect HCM if it is present, perhaps pointing to a need to screen at a young age before athletes start to train seriously.
If long term exercise induced adaptions do indeed worsen HCM or other SCD risk factors, there is a plausible mechanism whereby EPO might play a role, as it presumably allows for increased training intensity.
Finally, I think it's worth noting that any analysis comparing athlete death rate to non-athletes might give a very distorted picture. Surviving SCA usually depends on being lucky enough to have medical help on hand at the time, which is often the case for athletes. For example, Muamba is thankfully not a statistic, without that immediate medical help, he would have been.
Thanks for bringing this subject up again, I found it interesting and was frustrated when it got drowned out by emotive clamoring last time.