WIP Loosening lower B mass cut in B2XMuMu + using VeryLooseKsLL instead of LooseKsLL in S21r0p2

Loosening the cut will improve the control over the tails in the K*->K+pi0 channel. Has already been done in S28 (see !629 (merged) (diffs))

The StdVeryLooseKsLL container increases the number of KsLL candidates by roughly 50%.

Rates stay well below 0.5% (see below)

StrippingReport                                     INFO Event 10000, Good event 10000
 |                               *Decision name*|*Rate,%*|*Accepted*| *Mult*|*ms/evt*|
 |_StrippingGlobal_                             |  0.2300|        23|       |   8.710|
 |_StrippingSequenceStreamB2XMuMu_              |  0.2300|        23|       |   8.701|
 |!StrippingFullDSTDiMuonDiMuonHighMassLine     |  0.0100|         1|  1.000|   3.389|
 |!StrippingFullDSTDiMuonJpsi2MuMuTOSLine       |  0.1100|        11|  1.000|   0.217|
 |!StrippingFullDSTDiMuonJpsi2MuMuDetachedLine  |  0.0900|         9|  1.000|   0.039|
 |!StrippingFullDSTDiMuonPsi2MuMuDetachedLine   |  0.0300|         3|  1.000|   0.028|