WFPC2 coordinated parallels with ACS/WFC as prime:
Why does adding a WFPC2 coordinated parallel to an existing ACS/WFC result
in large changes?  I'm having trouble including WFPC2 parallels, please advise.

Reply:
I have received an answer from the RPS2 developers regarding the low
efficiency of adding WFPC2 coordinated parallels to ACS/WFC exposures.
The reply was that ACS buffer dumps and WFPC2 readouts (not shown on
RPS2 output) cannot occur at the same time, this led to one or more
of the ACS buffer dumps becoming "uncovered".

 The problem is therefore similar to the constraints that apply for 
ACS autoparallels, see pp. 162-167 of the ACS Inst Handbook (bottom of
p. 165 in particular).  The parallel observation needs to cover the 
primary ~340s buffer dump from the preceding WFC exposure, and still
have time left to complete its readout before the prime starts reading.

  Unfortunately there is nothing in RPS2 that will facilitate building
ACS+WFPC2 parallels that work in the sense that autoparallels do this.

  I did find a combination that provides an efficient result for use
of two filters in one visibility period:

Exp. 1:  2x520s ACS/WFC + 2x500s WFPC2  in one filter
Exp. 2:  2x520s    "    " 2x500S   "    in a second filter

This combination takes about 53 min including the guide star
acquisition, so it would often fit within single orbit visibilities.
[But if I recall right your particular visibility period was unusually
short and smaller than this.]

  I don't know the full range of parameter space for efficient results.
However, since for repeated exposures the parallel observation needs
to extend beyond the ~340s ACS buffer dump and still have time to readout
it may not be possible to drop to 400s combinations (I experimented some
without success).

  In the above specific example I was able to also add 2x500s with STIS
without any additional overhead.

  The key new information is that ACS buffer dumps and WFPC2 readouts
cannot overlap in time.  Thus the allowed range of effective coordinated
parallels between ACS/WFC and WFPC2 seems rather restricted.