Just based on length, it looks like skipping 22 will save the most distance. I scanned the map into OCAD and played around with the Skips and here is what I found:
Skipping 22, the course length is 11.3K
Skipping either 16, 17 or 18, the course length is 11.5K
Skpping 14 which you did, the course length is 11.7K.
Note the above doesn not factor in the difference in climb or ease of finding the control/route. Also, since I am working with a scanned version, the distances may be off slightly (less than 1%).
I'd still be under a 9 m/km pace, but I did save a lot of climb.
There is a split analysis here
http://www.attackpoint.org/raceskips.jsp/race_1127...
It looks like you would have been 3 minutes faster if you would have skipped 22 (gain 4:30) versus 14 (gain of 1:30).
It also looks like you had a few mistakes based on the splits. Not bad (3:30+/- total). It would be good for you to look at where the splits say you made mistakes and see if you think you did or not. If not, than you probably didn't have as good of a route as you could have.
Check out the Billygoat thread on AP and see the Route-Gadget-like video of the competitors who provided Garmin data. I wish you had yours to add, I'd like to see where you were in the pack as you went along. It's pretty interesting. Nice run for you, regardless. So much fun to see you out there repeatedly.
Sounds like you made the right decision. If 22 was the hardest control, you would be missing out of good training if you didnt take it.
22 was actually a fairly easy control. Very easy, in fact. It would have made more sense to skip it, I suppose.