Let me try to actually answer the question that you asked.
One of the things that Daniels took out of the 3rd edition was a table that calculates threshold pace for runs of longer than 20 minutes. a 3:30 marathoner has a VDOT of close to 45 and a marathon pace of 8:00. If you were to look at the second edition rather than the third, you would see that for a runner with a 45 VDOT, M pace and T pace for a one hour continuous run are almost identical.
This stems from the fact that lactate threshold is not a point of inflection, but rather a continuum. Waste products build up more slowly at this pace, but they build up more slowly. The point is to get a material accumulation of those waste products over time.
So yes, I would keep the long runs at M pace in. I would, however, avoid shorter runs done at M pace as they simply do not allow for a long enough accumulation of waste products to effectively effectuate the desired adaptation.