PV reconstruction efficiency in 2024-expected samples

The PV reconstruction efficiency is significantly lower in the 2024-expected samples than we were expecting (from memory, was around 95% on average previously) From the PrChecker:

All                  :  0.877 ( 22779/ 25988)
Isolated             :  0.920 ( 12280/ 13344)
Close                :  0.830 ( 10499/ 12644)
False rate           :  0.081 (  2015/ 24794)
Real false rate      :  0.081 (  2015/ 24794)
Clones               :  0.000 (     2/ 22779)

• The beamline looks like it is being read in the correct place (1,0.4)mm

• 50K events + conditions dumped using Moore v54r21 are available at /eos/lhcb/user/t/tevans/public/HLT1/minbias_nov2023/ (velo closed)

As an aside, the mu from this is 5.2, which is what it should be, which is reassuring!

@dovombru @mgiza @gligorov