Use tracking for projections in EcalVeto

[tvami]

Is your feature request related to a problem? Please describe.

Right now the EcalVeto uses the scoring plane to determine the recoil electron position, and thus the photon projection.

Describe the solution you'd like

To be more realistic we should use the extrapolation from the recoil tracking to do this.

Describe alternatives you've considered

For real data-taking this has to be done as there is no SP info in data.