Pare down the signals measured to just those relevant to the configuration needed

[EliotGann]

Too many data streams confuses analysis. If there was only on diode signal and if TEY was only measured when it might be useful, it seems like it might be a good idea.

1.) define new configurations with careful thinking and naming to be fully inclusive or possible modes of measurement, i.e. not just what motors should be where, but what signals are interesting to possibly measure. e.g. if I'm doing grazing incidence I want TEY and no diodes, if I'm doing WAXS, I only want the waxs beamstop, if I'm in downstream diode I only want that. maybe there are cases where TEY and a diode are needed.

2.) making those configurations (where the motors are)

3.) add the metadata to those configurations so that plans can pull a useful list of detectors and signals from the metadata. Not sure how that's going to work... a big lookuptable?