I did read that section about mortgages on ERN, and then I went back to the first part and opened a tab for future reading, and plan on rereading about the mortgage. We do have a mortgage, and it's right at $750k, so paying it off is not a minor choice, but up until now we had plan on keeping it, and possibly even recasting/refinancing it back out to 30 years just before FIRE, so that article has given me some more ideas to ruminate on.
One thing, about at least the ERN parts linked above that I read, that is made clear is that you kind of have to decide whether you are in the camp that believes that the CAPE10 continues to mean the same thing that it meant before the 2000s, or if you believe what some say regarding changes in accounting, etc. that translate into "naturally" higher CAPE10 values in recent years that can't necessarily be compared 1:1 with the more distant decades. Because what those charts all screamed at me more than anything was the difference between "all retirement dates" SWRs versus "CAPE>20" SWRs, even more so than which particular glidepath you use (and to make clear, the article stresses that using the glidepath, at least historically, in CAPE<20 retirement years provided NO benefit).
In other words, the analysis and commentary seems to suggest that if you believe the current CAPE>20 environment is going to result in sequence similar to the few scattered years around 1900, 1930, and 1965, then you should be planning an SWR of 3.5 and using a glidepath. But if you think that CAPEs of 20-30 are more equivalent to CAPE of 15-20 in past years, then maybe you don't need a bond tent/reverse glidepath and maybe and SWR of 4 still has a failure rate of around 5%.