[evlatests] More on Drop Behavior

[Barry Clark]

Also, when a script is first started, even if it is prepared by obs2script,
it starts as soon as it can get organized, and updates phases every ten
seconds thereafter.  At the first scan change, it reverts to the fixed
relationship with IAT.  I believe this explains all of Rick's "different
behaviors" of the dropouts, and does not further inform the ineluctable
fact of their occurance.

Have we checked, for the current firmware revision, that the dropouts
still occur with phase switching suppressed?  Could we be screwing that
up again?