An interesting subject. Wasn't it failed primary insruments in IMC that caused the 747 accident near Stansted a few years ago?
I suggest you shouldn't assume that failed gyros can only cause problems in IMC. Flying in poor VMC (very hazy, few k vis; 'fishbowl conditions'), you will use the instruments a lot to keep orientated, particularly if instrument qualified. Under these VMC conditions you may well be less alert about your checks, and it's easy for a vacuum failure to creep up on you without warning. This happened to my wife and I a while back, and it was very disorientating, despite being able to see something of the ground.
IO540 mentioned the autopilot. Mine derives its signal from the TC (it's a two axis type). Presumably you wouldn't want it to try to follow the heading bug on the failed DI, so it would have to be in wings-level mode or tracking a VOR or GPS.
It's important to cover the failed instruments, because it really is very hard indeed to ignore them.
Whilst I agree about the importance of keeping the partial-panel skills current, I would also enlist the help of the GPS. It is much easier to use the pseudo compass screen on the GPS than it is to use the conventional aircraft alcohol-filled compass, with all its errors - particularly if you have to fly a partial-panel approach. An IFR-certificated GPS is obviously preferable, but without DI and AI in IMC, you have an emergency on your hands and should use any tool that can help.
Similarly, if you lose your pitot/static system in IMC, the GPS will be of great help with altitude and groundspeed.
I have made a new year resolution to practice loss of gyros and pitot/static using the GPS to help.
AA.