SLB Nov Hackathon: Charm mesons builders tuning for SL

Investigating cut on the OWNPVVDRHO, corresponding to the radial distance from the best primary vertex.

The cut is 98% efficient for B->D(->Kpi)tau(->3pinu)nu MC sample truth-matched

Cutting at 0.2, seems to get rid of many multiple candidates in the event (plot taken from data block 7)

I also tightened the lower cut of the tau radial distance a bit (cc: @bfang).

BW TEST RESULT

Running on 100k spruce minbias sample, I got the following results in the SL stream.

master (GB/s)	sl-charm-builders (GB/s)	Improvement (%)
0.037	0.026	-30%

cc @groberts