I've had a few experiences that lead me to suspect this. A few times I've stored a small portion of wort for priming with gyle, and each time I re-boiled the wort for a short period, 10 minutes or so, which I perceived as producing a tremendous amount of DMS that wasn't in the fermenting wort. Also, a couple of times, I've had to re-start a boil, due to equipment issues or running out of propane. Each time the finished beer had a lot of DMS, even when the total boil time was over 2 hours. My suspicion from purely empirical evidence is that each time wort is heated after being cooled, it needs a full hour (or 90 minutes for pilsner) of rolling boil to drive off DMS, even if it had been previously boiled. Is this how DMS works? I had always assumed before, once you had boiled it off, it was gone, but that doesn't seem to be the case from experience.
Yes, DMS is formed every moment the wort is hot.
DMS is formed from heating of S-Methyl Methionine (SMM) that is found in all malts in different amounts.
Your boil burns off most of the DMS that was created during mash and preboil heating.
During boil DMS production is trumped by its evaporation, and it's 35-40 minute halflife depending on pH (why we have 90 min open boils, if it's allowed to condense and fall back in, you don't get rid of it.)
At flame out your wort will continue making DMS and DMSO until chilled. DMSO can't be boiled off as it's boil temp is much higher than water.
So.. If you have a stalled boil DMSO may be your problem.