In article ,
"Wyatt Emmerich" wrote:

I tried to fly under the hood relying on just the Garmin 196 artificial
panel. It's doable, but there were pretty big deviations and I had to
struggle. In real IFR in an emergency. . .I wouldn't want to be in that
situation. However, it is a nice reinforcer that helps me feel more
confident about my instruments in IMC.

I've tried that experiment once (actually, my student was flying; I was
just looking out the window and kibbitzing). I put us into some unusual
attitudes, and he did the recoveries. He did best if he completely
ignored the GPS pitch information and just used rudder inputs to achieve
zero rate of turn, relying on the elevator trim to take care of pitch.

The result was pretty ugly, but survivable. When he was trying to
control attitude with the synthetic AI and ASI, we just got into larger
and larger pitch oscillations and eventually I had to take control.