VTTs, Movement, and Maps
It's broadly known, at least in this particular arm of the RPG galaxy, that people have problems with VTTs. VTT stands for virtual tabletop, and generally refers to a category of software designed to facilitate running TTRPGs over the internet without losing much of the benefit of having an actual table. The problems that people identify are varied, but largely centre around things like the referee (DM, GM, Storyteller, etc.) losing control over where the PCs are going, them not supporting things like dungeon crawl procedures, the expectation for fancy battle maps and what that does to the shared imagination, and so on. My theory is that the problems people have with VTTs aren't so much a fault of the medium, but actually caused by Dungeons and Dragons. Original, I know.
I currently run a game in FoundryVTT, after being introduced to it by a previous referee/member of the current party. I like Foundry! I've had a good time with it. Before using it for my current campaign, I was a player in a game where it worked quite well, but it has certainly been different running this current campaign for it. The key difference is the playstyles that originate from the kind of system and adventure - the former was much more investigative in nature, where strolling round the map, talking to NPCs and fiddling with the environment was largely the point; combat was usually highly telegraphed, and traps and suchlike were pretty non-existent.
My current campaign is set in a dungeon. I'd hesitate to actively call it a dungeon crawl, largely because it's in Foundry. It's hard to run a "proper" crawl, with random encounters, checking for traps, attrition, and so on, when players are darting about the place. Monsters are harder to conceal, due to players having near total control of sight lines, doors are harder to make secret when there's an obvious gap or corridor to nowhere. I've pined for theatre of the mind more than once, and many referees do swear off VTTs due to issues like these, and it's easy enough to see why. On the other hand, I've had some pretty interesting moments, like a party order forming diegetically (that is to say, in-character) after one too many traps were set off and one too many fights were started with the squishiest characters in the front. I am also burdened, or possibly assisted, by the fact that what I'm running began life as a 5th edition first-party Wizards of the Coast D&D module, so things like random encounters just aren't really expected to happen. Which I think brings me nicely to my point.
The problem is actually Dungeons and Dragons - lack of dungeon procedure in 4th and 5th led to the problems we see today. 4th edition was when dungeon procedure went entirely missing from D&D books. If you open up a DMG (Dungeon Master's Guide) today, you won't find any meaningful advice for how to run a dungeon. 4th edition was published in 2008, and most major VTTs came out after this point. Roll20 released in 2012, Foundry was 2020, and so on. The only VTT still going from before 2008 that I can find (with my incredibly modest amount of research) is Fantasy Grounds. And what does that have? Token locking and referee approved player movement baked in. The simple fact is that other VTTs don't have support for a lot of the things people have problems with because it just wasn't asked of them when they were built; the pause button suffices. These VTTs can be extended with mods and the like to do what people want of them, but indeed that is just more work for the person running the game.
Of course, movement isn't the only thing that people complain about, but things like expectations around fancy and incredibly extensive maps are as much to do with the expectations of modern new players in general, people who have gotten into the hobby during this high-budget roleplay boom. Not something that is just unique to VTTs. I also, frankly, don't think that the issue of people taking maps, etc., at a fully literal face value is also the fault of VTTs, as in general people seem to struggle far more with the representational and symbolic nature of art than they used to. At my table, I just bring up Baudrillard every single time I have to remind them the map isn't a perfectly accurate representation of what's on it (JUST LIKE REAL MAPS), and after a certain point, it stuck. My description as referee is the ultimate truth.
It does still change the relationship with the map that RPGs have traditionally had, however, as often the map wasn't available to the players, and you weren't creating maps for every part of a campaign. It would be possible to create a more old-school mapping experience, again with mods or specific settings like zoom levels and persistent fog to create the capacity to get lost, but it's fiddly and burdensome. Furthermore, guess which major RPG publication all mapping procedures have slowly disappeared from? I can't find an exact date on this right now as I'm writing this at work, but it's definitely at least by 4th, probably earlier. This all being said, I've never had much of a problem with having to make the maps myself, because at least for my prep, being able to visually indicate so much on the map means I have fewer notes to write/peruse during play.
All this really brings us back to the point above about more work for the referee, which is often what VTTs can really cause, and really where the gripes lie. While some of it is expectations that can largely be ignored in the specific case, the simple fact is that it's harder to get a lot of VTTs to do what you want when it's not very basic rolling. I would be bothered by this less - I work in IT, and find screwing with this kind of thing fun! - but for a great many tables it's just not worth the investment. One of my earlier games over Discord just used Google Slides for battlemaps when needed, and it did more or less work, with interesting transitions to boot. You absolutely could also only use VTT maps for combat, as I'm sure some people must do, as it is the expectation that you map everything that's the problem, rather than the maps themselves.
In the end? Don't blame the play tool, blame the game, I guess. And some of the players. And definitely Matt Mercer (kidding).