In many ways, Civ IV was just "Civ III Conquests done right."
It was essentially the same game as III with one or two new gimmicks. If you read the designer notes by Soren, he doesn't really seem to have had any ambitions for the game other than to patch the macro design problems of III like ICS, then add a religion subsystem, and finally make the game moddable.
However, that by itself was enough to make a great game. Because after 5 years of playing III, a bunch of people, both designers and playtesters, understood the system really well and they came together to make a "version 2" that was just great. That's why IV is often called the best of the series. In many ways it's the culmination of 10 years of design, not 5 or 6.
Now compare with Civ VI.
Well, first, it's clear that VI is V with a lot of add-ons. There are so many GOOD, new ideas here. The civics tree, policy cards, districts, AI agendas. Some of the ideas, like Eureka boosts, are such a natural and perfect fit for this series that it makes you wonder why nobody invented them 15 years ago. There's a lot of great stuff.
The problem is that they built VI on top of V, and V (despite what people say about BNW) was never even close to being the equal of IV in terms of being an actual, deep, strategic game. So VI is plagued with all the problems V had - black box diplomacy (half fixed, to be fair), AI that can't handle 1UPT (clearly not fixed at all).
And this is doubly bewildering because the gap from V to VI was the LONGEST gap in the entire series. Lessons should have been learned from V and applied to VI, the way IV fixed III, but instead VI looks to be V unfixed but with a whole bunch of new stuff.