I think that there's a continuum, which is why I default to the position that good changes are good, and bad changes are bad. It is trite, obvious, and true. But there is no other way to put it.
Good changes are good and bad changes are bad.
What makes a change good or bad is ultimately arbitrary, subjective, and personal.
All changes - whether good or bad -
have a cost. They will disrupt games.
The designers need to acknowledge and embrace that, to consider that part of the price to pay for any cool new idea they might have, because that's the ultimate reality.
That's a conservative pressure, but that's what you get when you're working with an existing brand, with existing IP - you have a history to respect.
That doesn't stop someone with a Cool New Idea For RPGs from going out and publishing their own RPG with their Cool Idea, if they want, but it does mean that if you think it's cool for Dragonlance gnomes to be wild sorcerers, or for all tieflings to be from one nation named Bael Turath, that you should probably take into account the fact that some people will definitely not like that, and weigh that with the potential gains that you'd forsee from such a change.
I'd encourage the publishers to be about as cautious about reworking lore as they are about issuing errata or revising mechanics in 5e -
very. Do it, but know that it is going to disrupt some games when you do, and consider that as a counterweight to the value you see in it.