I'll try to make your claim more precise

I can think of a variety of ways in which drugs are more dangerous than training (some of which you mentioned):

• Drugs are easier in the sense that taking pills takes much less effort than hard training
• There are many more drugs (and they are more diverse) than there are forms of training
• Drug interaction is more complex and difficult to manage than mixing up types of training
• The variety of things in your body that you can break by drugs is much greater than the variety of things you can break by training
• Because of the point above, drugs are more likely to cause hidden damage which you are not aware of until it's too late
• Drugs are generally (but not always) faster acting giving you less time to detect a problem and correct it

I can probably produce more if I work at it :-)

If I understand your argument, it's that some athletes will have less of a safeguard, not that doctors are not much of a safeguard in general.

My point is that "someone with a medical degree is around" doesn't actually provide much safety by itself -- it all depends on the context. A wealthy US doctor (who can be sued, stripped of a license, kicked out of his country club, etc.) has a very different set of incentives than some guy with an M.D. in a poor country whose only chance of success in life is to extract superhuman performance out of a sports team he's advising.

the athletes would still be taking on unacceptable risk

Again, it's up to the athlete to decide which risk is acceptable and which is not. My concern is with a system of incentives, not with whether an individual athlete will make a "right" or "wrong" decision.

provided that the level of risk is equivalent for both

I think I'm more suspicious of long-term costs for short-term gains, rather than risk levels...