Maybe it’s gotten a ton better but a couple years back setting up and managing the whole stack had a “shitty on purpose so you pay for hosting” vibe. I noped out in a hurry—and I’ve built and managed some comms-related stuff that ought to be an order of magnitude more fiddly than just running a Matrix server.
Dealing with stuff like running a state compressor regularly, manually removing records it bails on. Also having fast enough resources to handle situations like someone joining the Matrix HQ channel (it's the biggest one) and pulling in like 70GB of database bloat. I hit a bug a month or two back where sliding-sync (the new protocol Element X uses) was causing huge amounts of database bloat, they fixed it but my instance was running something like 200GB on a server with less than 10 people.
EDIT: I run it on a hetzner dedi, I have run it on Apple fusion drives in the past and consumer SSD's and it was a terrible time. Need high IOPS so things don't grind to a halt on intensive db operations (like joining HQ channel lol)
I should have omitted the "I found it fairly easily to set it up with a domain of my own." part, because I do like the answers to the question, those are the sort of answers I was hoping for, so thank you.
the matrix server where i have my account, run by a small tech community with one admin recently had to switch from a blacklist to a whitelist approach in order to curb the amount of fake accounts and spam they were getting. as a side effect since that change i had to bother the admin multiple times because one of the groups i was participating in was not reachable. the amount of work the admin has in order to keep the server im shape is much more than i would be willing to tolerate.
on top of that not a month goes by where something doesn't break.
a security conscious friend who joined matrix because of me (well, he was interested before, but i finally gave him a reason)
just deleted his matrix clients in disgust after some messages randomly could not be decrypted.
i suspect that there is still a problem on the server i am using, therefore more potential work for its already overworked admin.