I wrote:

Have you seen the "HSI simulators" on modern handheld GPS systems? There
is a lot you can do with the data from a GPS, especially if you're just
trying to play back an instrument approach. The attitude can be estimated
fairly well from GPS data unless you're doing loops on the localizer.

I've been thinking about this more. It's not so clear to me that attitude
can be estimated well enough to use for evaluating a student's approach.

I'm thinking that the exact problems you're trying to identify (using pitch
vs. power for altitude control, etc.) are going to cause problems for
estimation.

I still think there's a lot of use for GPS-only data and that it's a whole
lot easier to collect than attitude info, but it's not likely to be
sufficient for all that you were wanting.

--kyler