All the talk about past expac trends you can just look at the
sub chart Blizzard themselves made last year. Legion was the last expansion before Classic came out and is the blueprint for the modern patch cadence with it being 77 days instead of 60 and ditching 3 .7 patches for a major .3 patch instead but you can still see the subs following the same trend of decline as usual. That had to be very sobering for them that it didn't matter how much effort they put into patch content but it brings up the question of why DF and apparently TWW have better retention with only a little faster cycle but lower quality and I think the boring answer is that the trading post probably keeps a chunk of people subbed all year even if they don't play often. Delves too probably contributed a lot but it's impossible to truly compare the numbers with Classic and how big China is now
I think Nymrohd is 100% right that the main problem is them making bad decisions from the start and having to spend dev time redoing them. Bfa's systems were heavily flawed from the start and they had to basically redo Azerite in a major patch and just gave up on warfronts altogether and I bet that contributed to why 8.2 felt so much bigger than either patch around it. You can even see it now with them already having to spend time rewriting quests because they did a poor job the first time around and now it's coming in the form of the addon apocalypse which is probably going to take months upon months to begin fixing since they're rushing it so fast and it'll inevitably harm content quality
The biggest tell for whether TWW was a fluke or not is how Midnight's minor patches will go IMO. If 12.0.5 and 12.0.7 look like this:
Tuskarr Feasts Revisited
Another Bfa island expedition map with daily quests
Horrific Visions Revisited Reforged
Dragonflight Timewalking
Then it's safe to say they prioritize content cadence over everything else and we're basically back to Bfa cadence but with reheated slop in between patches