You have to remember that prior to Falcon 9 there was no reference in what is "normal" for a propulsively landed rocket stage. Falcon 9 is now writing the initial "handbook" for this, including all the phenomena observed after touch-down.
I might be well off-base here, but that sort of sounds like the thinking that lead to the loss of Columbia? I don't think post-landing fires are quite the same as heat-shield damage, but imo "normal" should only be decided once Block V has proven at least 5 or 6 unrefurbished reflights.
Incorrect. Post-landing fires are a problem if they are of the type that does
unexpected damage. Like having an extended fire INSIDE the octoweb when there is not supposed to be one.
They have had those on early landings, but not anymore, courtesy of progressive improvements to prevent fires inside the octoweb. Short duration fires on the outside of the stage are now effectively remedied by the stage TPS and are expected. You don't do propulsive landings - of the type SpaceX does - and expect flaming stuff to NOT hit the legs or bounce back up against the stage itself.
Trouble was that SpaceX initially did not quite know what kind of a local environment a landing Falcon 9 would create. So, during the first several landings they ran into some surprises. Things they hadn't expected yet happened anyway. That's "writing the book" on this stuff. The way SpaceX operates they learned from their observations quickly AND next began to "harden" Falcon 9 to better withstand the landing-environment and subsequent results. One of the best observable effects was that eventually the post-landing fires INSIDE the octoweb went away. A substantial part of the effort being put into Block 5 is taking protection from the landing-environment to the ultimate level. So, SpaceX has been busy fixing the problem, and they are succeeding.
Columbia was completely different. That was normalization of deviance. In stead of fixing their problem (prevent foam-loss all together AND harden your heatshield to take on anything) they just comforted themselves expecting foam-loss, and other debris, to never be an issue because it had not killed in orbiter in 100 previous flights. This despite the fact that STS-27 was one heck of a wake-up call.
That's the difference: NASA failed to fix a problem despite getting warnings about the problem on almost every flight. SpaceX however got warnings of a problem and has acted (and is still acting) to fix the problem.
That doesn't go to say though that they are infallible. They had plenty of indications of a problem with their COPV's on multiple flights. They didn't begin to fix that problem however until after AMOS-6. They learned their lesson there. Unlike NASA, who knew - even before Challenger - that shuttle had several major design issues, but never bothered to fix them properly. Even the extensive improvements made after Columbia fixed only about half of the problem.