scott moore writes:

Its a good example of "fail safe stupid". Everyone had a
good reason for getting there, but the result is fairly
insipid. I'll probably "solve" it by permanently taping some
dark plastic film over the front of the instrument.

The behavior makes sense for IFR, but less so for VFR. If you are flying IFR,
you need to be able to see your instruments, and the effect this has on being
able to see outside isn't that important, since you may well be in IMC,
anyway. If you are flying VFR, you need to be able to see outside, but you'd
be less likely to use an autopilot in that case.

If you disengage the autopilot with the lights turned down, do the lights on
the autopilot then go dim?

There are a number of possible design choices here, none of them ideal. I
would not rush to second-guess the designers on the behavior they actually
chose without carefully considering all the possibilities and options they had
to consider themselves.

In any case, if the lighting on the panel really were to
fail, how having one bright "autopilot on" light in the
midst of a sea of dark gages would help escapes me right now.

It could keep you straight and level while you figure out the other problems.
You might even be able to use some of the light from the AP to see the other
gages, if you had a mirror (which you would, since mirrors are useful survival
equipment).

How would a dark autopilot in a sea of dark gages be more useful?

--
Transpose mxsmanic and gmail to reach me by e-mail.