Empty cruise
Surely you are not talking about correcting the outbound timing so as to achieve a 1 min. inbound leg???
The examples I gave for outbound timings were to illustrate the leg-length required to achieve a turn inbound at a point co-incident with the end of a 1-min outbound leg in still-air. I believe this is the best way of keeping the ground track of holding pattern as close to the still-air racetack pattern as possible. I think that 1 sec for every 2 knots HW is the best compromise between simplicity and accuracy.
I seem to remember that in basic training I was taught to commence timing from the latter of wings level/abeam radial and then note the elapsed time back over the beacon. If the time was greater than 3 mins, the error was halved and subtracted from the previous outbound timing. Similarly, if the time was less than 3 mins, the error was halved and added on. The aim was to achieve a consistent 3-min outbound/turn/inbound segment.
I agree that achieving a 4-minute hold doesn't really matter per se and the aim of the Mental Dead Reckoning is only to ensure that the aircraft stays within the protected area, when there is only an NDB available to fix postion. An FMS or GPS groundtrack achieves this much more accurately than any MDR process. Some larger/faster aircraft are limited to 25 degrees of bank in the turn and hence cannot achieve a Rate 1 turn. The holding pattern will actually take nearer to 5 mins due to longer turns, as is the case with my current aircraft. We don't bother to refine the timing using the previous error, and simply recalculate each time once steady outbound using 1 sec per 2 knots of HW - but that is the great advantage of having an instantaneous wind calculated by INS/FMS.