I am not defending the study protocol, but if performance changes in submaximal tests with (implicitly) lowish subjective RPE tell nothing about changes in ultimate performance capacity, the performance boost after the 135 ml should be equally dubious. From the second I saw the data, I had the impression that the 4.7 % boost from Hb (Hbmass) increase of only some 3 % was too much, which raised the question how well the blind-protocol worked.
Regardless of what measures to hide the group (test/placebo) takes place in the lab, it should be somewhat difficult to hide the real life consequences of donating 900 ml blood (Hbmass falls by some 15 %) particularly if the test subjects are recreational athletes with access to HR monitors, powermeters etc. with their Android apps either running as usual (placebo group) or bitching and whining about their performance every fifth second (blood doping group).
I don't know whom I want "more", but only that the magnitude and scope of all anti-doping policies should be based on rational basis and not on dubious science, moral panics or fearmongering.
And I wouldn't want dickpounds and travistygarts to make alone decisions in any of those things. Even when anti-doping activists are one voice in the public discussion, every bureaucracy tends to overestimate its relative importance and future capabilities if they just had the extra money. If we are to believe the anti-doping activists and scientists, the paradise lost free from doping has been just around the corner for the last forty years with a few extra million in their budget.