Require explicitly that the Velo state used in PrMatchNN is the one from EndVelo

Running over 1000 Events I got:

Old:

< MatchMCCheck         INFO **** Match 71560 tracks including          13903 ghosts [19.43 %], Event average 15.09 % ****
< MatchMCCheck         INFO   01_long                        : 53437 from    63190 [ 84.57 %]    333 clones [ 0.62 %], purity: 99.42 %, hitEff: 91.44 %
< MatchMCCheck         INFO   02_long>5GeV :   36382 from    40037 [ 90.87 %]    141 clones [ 0.39 %], purity: 99.59 %, hitEff: 93.39 %
< MatchMCCheck         INFO   03_long_strange                : 2564 from     3196 [ 80.23 %]      7 clones [ 0.27 %], purity: 99.08 %, hitEff: 90.00 %
< MatchMCCheck         INFO   04_long_strange>5GeV :    1404 from     1564 [ 89.77 %]      1 clones [ 0.07 %], purity: 99.45 %, hitEff: 93.14 %
< MatchMCCheck         INFO   05_long_fromB :      55 from       69 [ 79.71 %]      0 clones [ 0.00 %], purity: 99.30 %, hitEff: 91.06 %
< MatchMCCheck         INFO   06_long_fromB>5GeV :      41 from       47 [ 87.23 %]      0 clones [ 0.00 %], purity: 99.49 %, hitEff: 93.40 %
< MatchMCCheck         INFO   07_long_electrons              : 2774 from     4971 [ 55.80 %]     21 clones [ 0.75 %], purity: 97.62 %, hitEff: 89.42 %
---

New:

> MatchMCCheck         INFO **** Match 71948 tracks including          13625 ghosts [18.94 %], Event average 14.68 % ****
> MatchMCCheck         INFO   01_long                        : 54061 from    63190 [ 85.55 %]    331 clones [ 0.61 %], purity: 99.40 %, hitEff: 91.43 %
> MatchMCCheck         INFO   02_long>5GeV :   36729 from    40037 [ 91.74 %]    144 clones [ 0.39 %], purity: 99.59 %, hitEff: 93.38 %
> MatchMCCheck         INFO   03_long_strange                : 2590 from     3196 [ 81.04 %]      7 clones [ 0.27 %], purity: 99.06 %, hitEff: 89.98 %
> MatchMCCheck         INFO   04_long_strange>5GeV :    1413 from     1564 [ 90.35 %]      1 clones [ 0.07 %], purity: 99.44 %, hitEff: 93.13 %
> MatchMCCheck         INFO   05_long_fromB :      57 from       69 [ 82.61 %]      0 clones [ 0.00 %], purity: 99.26 %, hitEff: 91.37 %
> MatchMCCheck         INFO   06_long_fromB>5GeV :      42 from       47 [ 89.36 %]      0 clones [ 0.00 %], purity: 99.41 %, hitEff: 93.56 %
> MatchMCCheck         INFO   07_long_electrons              : 2812 from     4971 [ 56.57 %]     21 clones [ 0.74 %], purity: 97.60 %, hitEff: 89.33 % 

Looks like a nice improvement for free.