Fixes to lines targeting strange decays

Following-up on the results obtained after processing early data from 2025, we have spotted that there have been some issues in some lines targeting strange decays that needed to be fixed. In this merge request, we relax the impact parameter requirements of the tight proton selections which was killing most of the signal in cases like the selection of \Lambda\to p\pi^- candidates.

Changes in the rates (we report only those lines that experience changes)

	rate (before)	rate (after)	difference	difference (%)
Hlt2RD_Xi0ToPPi_Tight	33.8236	67.6472	33.8236	100
Hlt2RD_Lambda0ToPPi_DD_Tight	0	33.8236	33.8236	inf
Hlt2RD_Lambda0ToPMuNu_Tight	0	33.8236	33.8236	inf
Hlt2RD_Lambda0ToPENu_Tight	11.2745	33.8236	22.5491	200
Hlt2RD_Lambda0ToPMuNu_DD_Tight	11.2745	22.5491	11.2745	100
Hlt2RD_Lambda0ToPPi_Tight	0	11.2745	11.2745	inf
Hlt2RD_XiMinusToLambdaPi_Detached_Tight	0	11.2745	11.2745	inf
Hlt2RD_OmegaMinusToXiMinusGamma_Tight	0	11.2745	11.2745	inf

We followed these instructions assuming an input rate of 1355.471 kHz according to run 321707 in the RunDB. Numbers were computed with commit d43ccd0d.

The impact on the bandwidth is very small, and scales directly with the change in the rate (estimated to be <10\text{MB/s}).

We also implement a fix to avoid redundancies in a decay descriptor that computes the charge-conjugate of a self-cc decay.

This MR targets !4901 (merged).