Implement all update formulae and make them optionally usable

[ppebay]

This is in order to:

1. make LBAF have a behavior closer to that of vt
2. verify the formulae of the ACM-ICPP'25 paper
3. make this versatile by allowing to switch between "global information" computation of the transfer criterion (faster but less vt-esque) vs. the "limited information" one (more vt-esque but more complex).