I personally can't see the problem with using MDR. If applied correctly, the "clock code" method gives answers which are close enough to the "whizz wheel" method as makes no difference.
For example, the IR students I train use the clock code method to (very accurately) work out the drift to be applied in holds and on approach procedures.
If we are talking about visual navigation, then the student pilot should also be reading ground-to-map and map-to-ground and therefore the MDR is not the sole navigational reference used.
If it radio aids navigation, then the VOR/DME is used to fix position and again, the MDR is an aid to placing yourself close enough to the desired fix such that appropriate corrections may be made.
Let us also not forget that the 2000' wind used in the nav computer calculations is actually just a forecast - so even though the whizz wheel can calculate drift to 0.000001 of a degree, the information being fed in is just a guess by the met man.
If, upon reaching the first waypoint, the wind assumptions are found to be incorrect, then a correction may be made. If the trainee is used to calculating drift etc. using MDR then this airborne replanning is actually EASIER for him than for the student who relies solely upon the computer.