Modern math is esoteric to even the other sciences. Theoretical physics is probably its closest cousin. But if math has been bad for physics, physics has been great for math.

If some concept from modern number theory (>2000 AD) is involved in a core way in a commonly used software library for performance, then yes. Else no.

I will judge in good faith and if any issues arise, I will ask a random panel of 5 CS/math authorities for vote.

If this question were about >1900 number theory, or any >2000 math, or any applications of >2000 number theory at all, then it would be more interesting. The simultaneous restriction to a specific 30-year period (and the latest one at that), a specific field of math, and a specific application makes this very unlikely indeed.

Also, math hasn't "been bad for physics" lol, what are you smoking?