The data isn't subjective but the framing can be. The logic here would be that any man could be MSM and just isn't being honest about it. Banning all men would remove that potential high-risk completely. It doesn't make it a sound solution, however.
You're being extremely disingenuous when presenting this "logic" loop as some type of argument against baring high-risk populations from donating blood. The logic you're presenting is any group who has a higher risk than the next should be banned because a high-risk group is currently being banned from donating, and that you keep doing this endlessly, removing the next highest group until... what?
If your argument is that people could lie, then nobody will give blood because a women could easily lie about having sex with an active drug user, or lie about having sex with a bi-male who was sexually active with another male within a year.
What you're arguing is not how infectious disease risks function. Men aren't at higher risk because they are men, men are at higher risk because of the vectors of infection that disproportionately effect populations of men which are part of the screening process. Just like Black Americans aren't at higher risk because they are black, it's because mainly poverty is an excellent vector in aiding the spread of infectious diseases due to the lack of healthcare and proper government programs, and statistically the black population faces greater income inequality. So when you add in the effect of poverty, coupled with the risk of engaging in activities that have a higher risk of HIV infection, you get the outcome that black males in America are the largest HIV population by race.
Which is why when I posted about the screening process, I noted that it seems designed to be broad and wide and specifically designed to avoid minute details and breakdowns of higher risk populations. Could they have more specific screenings for populations that are high risk to deduce lower-risk sub populations? Very likely yes, but they aren't doing that.
Now my question is why aren't they doing more detailed screenings? Is it an issue with statistical accuracy, cost, money, risk and so forth?