The article identifies a very specific problem. Women will swipe through the app and reject the bottom 95% of the men immediately, but then all of them women end up matching with only 5% of the men. This is useless to both the 95% of men who don't get any matches and the 95% of women who can't all be in a monogamous relationship with that 5% of men.
In theory you could just stop showing the top 5% of men to anyone, but that's pretty useless. Then the 90-95th percentile men would get all the matches and you haven't solved anything.
You could assign percentiles at random so that some 80th percentile women only see 40th percentile men and vice versa, but that seems less likely to work for obvious reasons.
The problem you need to solve is to get the 60th percentile women to go on a date with the 60th percentile men to begin with. What happens then is up to them, but at least you're putting people in a situation that could lead to something, instead of the one:many matching that sucks for everyone except the players.
Again, this is all predicated on what data can be measured inside the app.
The top 5% of men would be a completely different group if it was measured differently. It's not the actual men we are ranking here, it's their profiles.
In theory you could just stop showing the top 5% of men to anyone, but that's pretty useless. Then the 90-95th percentile men would get all the matches and you haven't solved anything.
You could assign percentiles at random so that some 80th percentile women only see 40th percentile men and vice versa, but that seems less likely to work for obvious reasons.
The problem you need to solve is to get the 60th percentile women to go on a date with the 60th percentile men to begin with. What happens then is up to them, but at least you're putting people in a situation that could lead to something, instead of the one:many matching that sucks for everyone except the players.