Personally, I don't see that as Conservative (if you consider that word to be the opposite of Bold or Experimental). 5e has done a lot of bold experiments. Even producing less books was a bold experiment. Not everyone agrees with "produce less, sell more" - it's had to be proven.
This is the crux of a lot of the disagreement here, I think, because we have two different, yet somewhat related things being discussed.
The first is the marketing of and general approach to selling/providing settings, which I agree, has not been conservative in that sense, but rather has been pretty bold, and yes, has taken some significant risks.
The second is the design and nature of the specific settings themselves. That, I would argue, has been extremely conservative. Very few risks have been taken. That's not necessarily a bad thing. You don't want to lose lots of money or worse. But it is a thing. And we can't say something isn't conservative in this sense, just because we think its a good plan (I mean, hell, we've all seen a lot of good plans in D&D which were very conservative compared to what other party members proposed!).
SCAG - A setting book focused on the most broadly-known and arguably most popular part of the Forgotten Realms campaign setting, which is certainly the most broadly-known campaign setting at this point.
Ravnica - A setting book focused on a fairly popular MtG setting, and one which in many ways reuses concepts proven popular before in D&D, but in safer ways (Planescape and Eberron come particularly to mind). This is, imho, by far the riskiest setting book they've done for 5E. But I would say it was not very risky. The risk was carefully mitigated by the fact that the only other official setting book at this point was SCAG and that was years before, and the fact that they'd tested the waters with Planeshift stuff. Risk is also reduced by not needing to commission much art.
Acquisitions Incorporated - Very low-risk, keying off a captive audience for an enduringly popular podcast from long-popular cartoonists. Nice merchandised tie-in that must have had significant minimum sales expectations.
Eberron - The most popular setting (I would argue) of 3.XE, which remained popular in 4E, so very low-risk, known audience and so on, and even then, they very carefully tested the waters with a digital predecessor with limited material, to see how that would do, before creating a full-size setting book. Heavy art re-use also reduces risk.
Wildemount - Setting from a staggeringly popular and broadly-known podcast/stream (whatever you want to call it), which is popular it probably significantly increased the sales of D&D, and which had previously achieved good sales with their own setting-book despite not having the reach/resources of WotC. This is no-risk, no-brainer stuff. It's a largely generic and inoffensive setting, and like all 5E settings, kitchen-sink in design.
Theros - Too early to call in some ways, but for whatever reasons, Ancient Greece/Mythic-style settings are kinda zeitgeist-y right now (I feel like God of War kind of kickstarted this, quite a few years ago, but its certainly rolling), with like half-a-dozen 3PP ones (or related ones), some of which are clearly doing really well. Combine that with the success of Ravnica (proving MtG settings can make it) and the fact that you don't need to pay new artists as much, because you can re-use art, and you're riding high.
So the question I guess is, relating back to the OP, is this more conservative in setting design than any previous edition?
1E - Clearly less conservative, but a very different situation.
2E - Ahahaha yeah no contest, I don't know if any game in history ever has or ever will have such and wild and crazy bunch of settings as 2E. From beginning to the end. It's fair to say 2E went too far here, but in the process, a lot of
truly beautiful things were made.
3E - I feel like 3E was a not a huge amount less conservative than 5E, in terms of setting-design, but it took a different approach, instead focusing on "what is popular with existing D&D players?", for the most part. Still, it produced one completely original, and certainly somewhat risk-taking setting, in the form of Eberron, which combined with the rest of 3E means it is less conservative than 5E has been so far (maybe the next two years will change that).
4E - The massive changes to cosmology, the Nentir Vale approach, the revival of a 2E setting which 3E skipped, and so on, to me, show less conservatism in setting design than 3E. 4E was a bit weird in that the default setting was a lot more heavily implied than other editions, too, much more directly tied to character abilities. Certainly this was bold.
So yeah, if we're talking about conservatism in terms of the design and type of settings being released, I think it's fair to say 5E is the most conservative, and by a noticeable but not massive distance. 3E isn't that far behind it, and 5E has years to run yet.
My prediction, optimistically, would be that by the time 5E is "done", and we're in 6E, it'll be the second only to 1E and 2E in terms of riskier settings having come out. I say this for two reasons:
1) I figure 5E has a lot of life left in it yet. It may well be the longest-lived of all editions of D&D. I could easily see 5-7 years from now before we see a 6E even being contemplated. And I think technology may well be the push, not rules stuff, but that's a whole other discussion.
2) Looking at a list of 5E products, it does look like 5E is becoming increasingly confident that for any even half-sane product, there is a sizeable and growing audience. Unlike 2E, where the audience was not really necessarily growing much (instead being partially consumed by other RPGs and card games and so on), 5E's audience has expanded massively, and whilst it may slow, I think it'll be a long while before it stops. We shall see of course with the post-COVID economy and so on, but still.