This is fine and acceptable as long as they are able to figure out the exact amount of testosterone the elite-level female athlete has who is not androgenous.
Then, they would have to titrate the substances in carefully in order to maintain that level.
Any less than that and they are cheating Semenya.
If no titration or monitoring policies are in place, then they should not rule in favor of this until they have an "A" standard for the testosterone levels of most female athletes. In other words, they would have to delve into major statistical analysis to be correct--it's not that simple.
Scientifically speaking, they would have to get a mean, median and, probably, more importantly, a mode for the female athlete. The mean will tell them what the trend is overall. The median will determine a direct midpoint for the levels, and the mode will let them know what most of the female athletes are doing. You would then need a control group of non-athletic females to see what their levels are.
You would need to test both of them in a workout situation to find out the maximum (or near maximum) they could reach in a situation that amplifies their testosterone levels (since they can switch situationally). You would then need to find the standard deviation because you would have to know the full extent of the ranges of possibility.
Only then could you create a baseline by which all female athletes (androgenous or true XX) should be at in order to state what is proper for an athlete like Caster Semenya.
Sadly enough, unless all of these are in place and the full opinions and adjudication are conducted in a fully clinical manner without bias, any ruling seems premature since the scope of androgenous levels may not be completely evaluated and understood.
tl;dr version: No, this is not good as it's not a fair rendering.