Fascinating!
My biggest question about this (and my biggest qualm with 'front-load your efforts' strategies thus far), though, is - how realistic is it, really, to expect 100% accuracy on the first ~25 questions? It's a simple enough strategy for a test prep instructor, or for a student who's going to end up with a very high Quant score anyways, but will it actually increase scores for people who are weaker on Quant?
I'm imagining a student of mine who's about at the '650 level' on Quant, meaning that he or she can realistically answer a bit more than half of all 650 level Quant questions correctly in two minutes each. If that student invested almost the entire time on the first 25 questions, he or she would have a bit less than 3 minutes per question, on average. Is that enough for the student to go from 60% accuracy at the 650 level, up to 100% accuracy at the 800 level? My instincts say no - my typical student probably wouldn't be able to sustain that performance through the first 25 questions, even with three minutes per question, given that the questions they'd be seeing would be much harder.
I do think that most GMAT questions are possible for most students to answer in
unlimited time, assuming they have the basic background and terminology they need. But in 3 minutes? That's not
that much easier than 2 minutes. And there will always be a small but significant number of questions that you aren't going to get right no matter what you do... and there'll be way more of those popping up with this approach. Thoughts?