// 0 - Use original function; 1 - Use extension implementation
// Multiplier of commons collision force
If you are curious what was the problem then keep reading.
You probably know about problem where with high tickrate and low nb_update_frequency value, common infected starts getting stuck at every possible corner.
It's because due to lower interval between updates common infected moves less distance and that starts triggering early return for collision check without a chance to escape collision resolution or try to solve it (See code references).
Yep that's it.
Why thread has "Experimental" in parentheses?
Because even tho I removed those 2 checks and this fixed the problem I still have no idea why someone added them and what could potentially break without them.