Adapt to new, consistent Enums for all Track versions and drop unused entries. Change to return by value instead of unique pointer in KalmanFilterTool.
The LHCbSequenceTimerTool that is removed from PrVeloUT draws a random number from the RndmGenSvc on initialization ( and then does nothing which is why it is removed). In retina mode, before the velo tracking, a special clustering is run, but even before that the VPRetinaSPmixer is run, that also draws a random number from the same service. Now removing the TimerTool seems to cause the Mixer to draw a different random number causing the Moore Retina tests to fail (fyi @rmatev)