If you only had resources to take blood from 100 people, would you choose:
Group A - 100 people who say they are straight
Broup B - 100 people including 20% who say they are gay
Keep in mind that you have valid statistics they show that a higher % of homosexual donations fail the screenings.
The answer is obviously Group A. Of course if the resources could be procured, both groups should be used.
I used a purposely ridiculous solution to show how hard and ridiculous it would be to guarantee blood would be disease free to save money from doing the tests. Surely if money is all that matters, making a population with 0% risk of having the diseases are even more preferable than the standard population. And if you want to ensure that you have a group like that, you have to go to extreme lengths to do so.
If I had that choice I'd simply flip a coin or something similar. Why? Well several reasons. The first one is simply lack of information. Only the sexuality of the groups is given, but you're not telling me how many (if any) are drug users. Also there's no information on how many pratice safe sex. And to finish up, there's the simple thing in geoup A's statement. that they 'say' they are straight. They could simply lie. As such there could be just as many high risk people in group A and the people that are classed as high risk in group B could be taking all the precuations and be perfectly clean.
And, of cause, as with any statistic, in practice it becomes rather meaningless. Gay people may be more likely to have those disease, but that doesn't instantly mean that any random group is going to abide by the statistics. By the same token just because straight people are a 'lower risk' does not mean that those in either group will have less instances of the diseases showing.
So like I said, I'd be as happy choosing either group and I'd not make a snap judgement (as you have) unless I had more facts. Like you say, the ideal is doing both groups if you have the resources.