I suspect that it would have the frothing reactionaries out in force, which is the main reason it wouldn’t be tried; but it would be interesting to know, if such cultural considerations were not an issue, what sort of structure one would use along those lines:
You couldn’t have too many distinct bands, since each one would be another medal category, another competitor or team participating countries would have to field; and there are finite reserves of both audience patience and competitor resources; but you would want enough of them that people are generally competing with not-too-markedly-dissimilar competitors.
I would be curious to know how many bands that would end up being(and whether there would be additional complexity imposed by team sports medicine and the endocrinologists identifying other relevant factors, so that a ‘band’ would actually have to be a somewhat irregularly shaped region in a multidimensional space of assorted parameters):
Among sports that actually do something like this, boxing seems to be on the high end(currently 10 different weight classes, I think); I don’t know if that reflects a judgement on amounts of extra weight that just aren’t important enough to bother with; the need to keep the bands wide enough that a given athlete doesn’t change weight classes every other meal; or if they’d actually like to have 25 classes but the logistics are problematic.
As for “corralling genders off away from each other”, in a lot of sports that basically depends on whether you want to have any female medalists or not. The 800m looks like it would be one of the ‘or not’ sports. For the 2012 and 2016 800m(which were the ones Semenya competed in); 2012 women’s times ranged from 1:57.23 to 2:00.19 (after a couple of russian disqualifications, one of which would have been 1:56.19); men’s times ranged from 1:40.91 to 1:43.77. in 2016, women were between 1:55.28 and 1:59.57, men between 1:42.15 and 1:46.15.
I’m not quite sure why the IOC has such an interest in Semenya’s situation(she was only .3 seconds faster than the non-disqualified russian; and slightly over a second slower than the disqualified one in 2012; and 1.2 seconds faster than the next finisher in 2016) specifically: they can’t exactly be surprised that Olympic level athletes are a roundup of assorted wildly atypical performance characteristics; and I’m unclear on why they seem so interested in drawing the line there(especially when there is still such a delta between the men’s and women’s results overall: Why would you get a ‘not-woman’ designation for a time that is 1.2 seconds ahead of the next woman; but 13.13 seconds behind the male medalist?)