Draft: ParKF: Gives velo chi2/ndof and tracking qop to states

Proposed work around related to Issue 590 about problems with the single high pt lines an Z-> ll samples.

Attach the ParKF chi2 of the Velo only to the states as well as the tracking estimate of momentum. Lines observing the described issues (see link above) can then use this instead of the full ParKF estimate. There will be no improvement to the chi2 values, just the option for line to fall back to this estimate. Also this is not the exact same chi2 as the velo-only (or current) chi2 estimate, as this uses slightly different parameters, e.g. doesn't account for residual momentum in VP.

NB: If the ParKF is not being used or the issue linked above is resolved this MR should be reverted, since it would just save duplicate / unused data. (if the Velo_only Kalman sequence runs the entries are duplicated.)