Every time this topic comes up, several people frame the issue as a choice between:
A. Following the obvious financial conclusion, or
B. Sleeping better at night.
It's not that simple.
First, the financial conclusion is not so obvious. It all depends on the future, which by definition is impossible to predict with any certainty. What if the US experiences 20 years of Japan-style deflation? What if the markets go sideways for the next 10 years? What if your interest rate is 5.25%, your balance is too low, and term too short, to justify refinancing? Each case is different. You can plug your mortgage numbers into FIRECalc to quantify the potential success rate of holding a mortgage; and while it may be high under certain parameters, it is certainly not 100%. And... there are lots of valid reasons (discussed in previous threads) why the FIRECalc history may be overstating success rates on the hold-a-mortgage decision analysis.
More importantly, it seems to me that a heavily-leveraged balance sheet in retirement is simply an inappropriate capital structure for that stage of life. Especially if you've already got the ways and means to support your desired lifestyle without taking that risk. I'm now in a mode of minimizing uncertainty and risk of failure, not maximizing net worth. If you are COUNTING on the market outperforming your mortgage to enable ER, that's even worse. OTOH, if you are inherently risk averse, and just want to build a bigger legacy for your kids, then go for it. The odds are decent that you'll come out a couple points ahead, especially if you have a really low rate and a long term remaining.
But don't frame this decision as thinkers versus sleepers. I'm the former, not the latter, and I happily paid off my mortgage at ER using funds from my taxable account, and the rationale registered above.