Whilst I cannot add anything to the above replies as I have never flown a "glass cockpit" anything there is one issue about the older "steam driven" stuff that concerns me.
The increasing regulatory demands for new equipment in old cockpits is leading to some very unsatisfactory layouts. On a flight a few months ago in IMC that required an initial NDB approach with a subsequent missed approach (could not get visual by the missed approach point) and then an ILS into a nearby airport I was struggling because of the cockpit layout where some information was well outside the scan pattern.
At present in Australia there is an on-going investigation into this country's worst air crash for some years. The aircraft, a Metro 3 (or 23) was conducting a GPS approach into a north Queensland aerodrome when it hit a ridge in IMC. The investigation is being hampered by an inoperative CVR and a damaged FDR. Without wishing to be seen to pre-empting the findings (which I suspect are going to be inconclusive) the aircraft appears to have been in a higher than normal rate of descent which presupposses a loss of situational awareness by the PF. There has been a lot of discussion in Australia about the way in which distance information is presented during a GPS approach and that it is not always showing the distance to the runway.
The GPS was a later inclusion in the cockpit of this Metro in which there is not a lot of room for the inclusion of anything. I cannot say where it was located but I suspect it was not easy view of the PF (this particular Metro was not equipped with autopilot).
I, for one, shall be very interested to see whether the cockpit layout in the Metro is considered as one of the factors in the crash.