Inconsistent handling of nominal drift speed in config

[brettviren]

The nominal drift speed as multiple avenues to be set which easily leads to inconsistency.

The main issue is that it may be set in the Jsonnet config and in the field response file (fr.speed). The latter typically "bakes" in a value when the field response file is transformed from whatever data came from the external field calculator (garfield, pochoir).

• Drifter parameter drift_speed set from Jsonnet.
• DepoTransform parameter drift_speed set from Jsonnet.
• OmnibusSigProc uses fr.speed.
• DepoFluxSplat uses fr.speed but allows overriding with drift_speed set from Jsonnet

There are others (in both WCT and larwirecell).

I think the best way to bring these into consistency is to adopt the approach of DepoFluxSplat (default from fr.speed but allow override).