John (and Matt) have forgotten more about this race than I'll ever know. I figured that I'd throw in my 2 cents as the greenhorn creator of that pace card, which is based on the splits they've described. Plan versus reality, in other words.
I targeted a 3:15 ascent for the front half of the marathon last year, and I got there in about 3:17. (3:17:43, according to the results page.) I had little in the way of altitude or high altitude training: one "ultra" 28 mile marathon up to 10,000 feet in Switzerland, and a few weeks of sleeping in an altitude tent at a simulated altitude of 14,000 ft. My climb training was mostly 8-10 mile treadmill runs at 15%, with one treadmill simulation. Trail training was some Virginia trail races, which have trails about as gnarly, or a worse, than Barr.
I was a little bit ahead of the pace card until Barr Camp or so, and a little bit behind it the rest of the way up. The altitude is definitely a beast, and unfortunately we don't know what our response will be until we're in it. There are people with high V02 max values that get clobbered much worse than others. From everything I've read (and I've pulled a lot of the research on running and altitude trying to figure out if that altitude tent is nothing but an expensive placebo) the reaction to altitude is highly individual, and variable even then.
I think it's definitely much better to be conservative in the front half. Whenever you get stuck in a conga line and feel yourself getting frustrated, instead turn it around and be thankful that you're having a conservative pace imposed on you for a while. (This, at least, is how I've learned to cope.)
There's plenty of time and space further up if you've really got the extra. I found out, I didn't have that extra.
