By most measures it’s an inferior tool. It lacks many of the advanced features of Foundry, and isn’t as simple or clean as Owlbear.rodeo - all of which actually appeals to me as my feature needs are low, but slightly higher than Owlbear.rodeo. Plus, I have a very large library of resources that I’ve specifically made easy to use in it and I’m far too lazy to put effort into converting.
The highlighted sentence is an excellent reason to stay with a VTT. Even if there is no lock in, it still requires a significant effort to migrate you content and learn a new tool.
Also I think Roll20 is a pretty darn good VTT and that is based on my experience before the recent upgrades. I didn't go with Roll20 mainly because even at the highest tier I couldn't store enough data for a single campaign I was running. But used it as a player fairly often for years, especially for one shots and on-line game conventions. For most players, it just works and a DM can pretty quickly explain to players what to click on. The learning curve, like most VTTs, is mostly on the DM's side. Foundry is pretty easy for me to bring new players into, but I've never had to step people through how to enable hardware acceleration in Chrome, like I have to with Foundry.
Also, as a player I've had bad experiences with games where the DM was running it from their computer, rather than a hosted service, especially if we were in different countries. Even on poor connections in countries with poor or highly-controlled Internet infrastructures, I've never had an issue with Roll20. Of course, like any VTT, huge maps with all the bells and whistles, can affect performance, but Roll20 has always been one of the most stable VTTs in my experience.
Also, I really love Roll20's find-a-game site. If you are looking for players, that is definitely an advantage.
I love what you can do with Foundry, but I spend a stupid amount of time testing and configuring mods, creating custom landing pages, doing fun things with maps. I really is a hobby in and of itself.
