Quite a while back, I posted some studies of EPO that showed that different individuals (not elite athletes, FWIW) exhibited different increases in HT to the same dose. So beyond the fact that a lower natural HT allowed a greater increase in the 50% era (and even after that, allowed a larger % increase--e.g., going from 40 to 42 is a 5% increase, whereas going from 48 to 50 is about 4%), some people just get more of a boost from a given dose than others. Then, on top of this, as thehog noted in his post about Vaughters, there tends to be a negative correlation between natural HT and events downstream that convert oxygen intake into power. So even if two riders get the same boost from EPO, the one with the lower HT tends to get greater oxygen utiliization and power increase. Only a rough correlation, but the point is, these are the kinds of things that can make big differences in how much power increase two riders on the same program might get. Even if you skip the reticulocyte stage, and just transfuse without EPO to raise your HT, there can still be differences in the increase in oxygen delivery to tissues and the resulting power increase.
Wrt warning riders about suspicious blood values, remember that the precursor of the biological passport was the off-score, and it was standard practice to warn riders when their off-scores were high. One of the most notable examples of this was Tyler in 2004. He failed the transfusion test in the Olympics that year, but because the B sample was messed up, didn't get sanctioned. Then in the Vuelta, he failed again, and this time did get banned. But before all that, back in April or May, he had a suspicious off-score, and was called in by UCI to talk about that.
You can certainly look at the off-score, then the passport, as a way for UCI to control doping rather than eradicate it. In fact, since the passport depends on a significant increase from a baseline, it's always going to be possible to blood dope to some degree under the radar. Again, I posted some links a while back to Ashenden's studies demonstrating this. But just a warning--"while we can't prove you're transfusing, we think you are, and we're watching you very closely"--has to be considered a worthwhile deterrent.