Weak and no-canon approaches are viable options, but they need to be weighed against whether they add or detract from the possibilities and potential of the target. Done well, and it's a set of soft reboots that enhance and update. Done poorly, and it goes to pieces fast. https://twitter.com/io9/status/1245445692736393218

For example: DC's serial massive reboots are, IMHO, an example of how not to do it, and how a company can fail to learn from its past mistakes. Continuity falls apart, they reboot, then do the same thing all over again. This contributes to a... https://comicsalliance.com/dc-comics-reboot-history/

..."Fractured Fandom" as well, where different parts of the fandom encounter different continuities (often at the same time and from different times), become attached, then encounter the "current" continuity and have a huge WTF?!? moment. A prime example of which was people...

...who encountered (in cartoons) Starfire in Teen Titans and John Steward in JL and JLU, then hit the comics to find a lacklustre overly sexualized Starfire and a bizarre love of Hal Jordan with John Stewart being shuffled off to the side.

In the #DnD side of things, it gets messy and complex fast. TSR's continuity was limited, at best; and Wizards have increasingly applied a loose or new canon approach. The issue here is the same one that DC ran into. Changes to one area have massive effects on others that...

...have to be accounted for. An example here is the rewrite of Elf Lore in MTF. Taken by itself, it's fine. however, it radically changes the histories and conflicts of multiple D&D worlds, not the least of which being the current baseline, Forgotten Realms.

This is where I make the argument that you're in for a penny, you're in for a pound. If you're going to do something that fundamentally changes everything, you don't get to hand wave it and say "Meh, we'll let the DMs at the tables sort it out."

That's unfair because not everyone has time to sit down and hash that all out.

Then there's the "value added" question. Does your loose or no-canon approach add value to the setting or does it detract?

From Chult and #ToA, we can see both. One one hand, the loose approach let them add more areas to the map and dip their toes into giving Chult a deeper...

...history than previously given. That was good. Value added. Except that the other places were all ruins and the "plot threads" (potential for further gaming there after the campaign) were absent or of such poor quality it detracted value. Probably the best example of value...

...lost was the Jungle Dwarves. Chult originally had two Dwarf groups in it, but they were mashed together into one for ToA and were net value lost. For the full details, watch my video on them over on YouTube. But the point is that the loose canon approach here failed.

Instead of adding value, it detracted it. Same with the MTF changes to Elf Lore. On the surface, value added, but dig down a bit, and without the supports to make it all work, it just created a mess of plotholes and continuity errors. Which is fine, as long as you only play in...

..."The Now" and never dig back into the lore. But people love lore, they love knowing what's going on, and what shaped things, so it's a short term win before people start trying to reconcile the new canon with what they can find for the rest of the world.

For an example of a loose/no-canon approach that worked, see the SCP site. Then bear in mind that it only works because of the literal nature of the SCP Earth/Universe where history and time are malleable & reshapable, and observable from places where things haven't happened yet.