I start from your last statement :
endurance athletes will continue to have a great incentive to do it. In particular when you are poor and have nothing to lose and you know the faster runners around you are doping.
This is true, but we need to look at the premise. And the premise is that THE COMMON FEELING, supported by Antidoping too, IS THAT EPO CAN GIVE A BIG ADVANTAGE (don't forget that scientists, not knowing anything about the training effects, spoke of 30" for 5000m, and the same in percentage (1 min in 10000, 2 min in HM and 4 min in full Marathon), all hypothesis clearly completely ABSURD for coaches used training athletes in clean way, that know where ALREADY was possible to arrive without any doping.
For example, I had athletes running under 27' in 10000m many times (Nicholas Kemboi 26'30", Ahmed Hassan 26'38", Imane Merga 26'48", Moses Mosop 26'49", John Korir and Mark Bett 26'52", Geoffrey Kirui 26'55") and in all the group only 2 can be considered as REAL CHAMPIONS : Nicholas Kemboi and Moses Mosop, all thye others were good athletes who found races well paced and achieved their PB remaining in the group.
At the moment, already 85 athletes ran under 27' : this means that that time is a NORMAL PERFORMANCE for GOOD TALENTED ATHLETES, but not real champions. In other words, Bekele and Gebrselassie were different subjects.
We live in a world where every day the economical power of BIG PHARMAS increases. In TV every 3 publicitary spots, one is for recommanding the assumption of some supplement, for different reasons (health, beauty, improvement in the efficiency, and so and so), so people are bombed by the idea that every kind of performance (not only in the sport) that is out of the average MUST BE DUE to some esternal help.
This creates the basic mentality for believing in doping : so the answer to the question "if EPO doesn't work, why a lot of athletes use : are all stupid ?", is "Are not stupid, but PLAGIARIZED by the continue talk about the advantages of blood doping.
We can discuss a lot about different training methodologies, but always having clear in our mind that training for a top performance is the correct mix between tough workouts and recovery. This means that making recovery faster is not important, because depends on the typology of the athlete and his technical choice.
Is it better to train at 80% of max level 3 times per week, with only one recovery day in between, or sometime at 90% recovering 2 days in between ? And what if we have some session at 100% of the effort, recovering in between 4 days ?
The chain, according people believing in the effects of blood doping, is : with the assumption of EPO athlete can recover faster, with more recovery can train more, and training more can improve his performances.
But the question is : DID THIS ATHLETE TRIED TO INCREASE IN CLEAN WAY THE SPECIFIC INTENSITY OF HIS TRAINING, USING MORE DAYS FOR RECOVERING, changing the classic schedule of "simmetric training" that doesn't allow to increase the level of one single session (otherwise he goes in overtraining), or never tried BEFORE using EPO ? Because, if never tried, how can compare what did before doping with what did after doping ? The limiting factor was in the mentality, because everybody understand that at the base of every improvement there is the increase of training.