For some reason, I was pondering over the problem (with really no background in mesh/geometry algorithms) - it feels very NP-complete-like problem. It just feel

For some reason, I was pondering over the problem (with really no background in mesh/geometry algorithms) - it feels very NP-complete-like problem. It just feels hard. Could you expand on some of the details of the algorithm in terms of how these 30 algos scale? Do they scale in polynomial time or do these algos scale in non-polynomial ways but the problem space is small enough for it to compute the solution quickly. Perhaps some epoch threshold has reached and the algos bail out?