Just to provide one example, if Firefox suddenly no longer has bookmark/history/password syncing because Mozilla has refocused on its core products (Firefox/Thunderbird/MDN), suddenly you'll see Firefox's market share dwindle even more, because ordinary users are accustomed to every browser having a bunch of bells and whistles like profile syncing.
The set of features people expect from a modern browser is really big now. To their credit, the Mozilla web standards people actively fought against a lot of the scope creep like "webpages should be able to flash firmware to USB devices" or "webpages should be able to talk to MIDI keyboards" but they lost, and now those are things a web browser is expected to do.
I've seen people argue that Mozilla shouldn't be offering cloud services and should just build a browser that never phones home to any servers at all, whether it's telemetry, automated updates, or profiles. I think all of those are part of shipping a modern browser, personally.
Maintaining all those cloud services raises your company's operational costs a lot, you now need people on-call 24/7 to maintain everything, you need webdevs who can wrangle postgres or redis or whatever, you need security experts to make sure the cloud stack is secure end to end, etc. So I think it's also fair for people to call this cloud stuff out as a cost center for Mozilla.
Which cloud services? I don't see how automated updates nor sync count as "cloud services" (and I'll note that the sync server used to be open source, so you could pull from the community like Mozilla claims to be part of).
The set of features people expect from a modern browser is really big now. To their credit, the Mozilla web standards people actively fought against a lot of the scope creep like "webpages should be able to flash firmware to USB devices" or "webpages should be able to talk to MIDI keyboards" but they lost, and now those are things a web browser is expected to do.
Keeping up with all the scope creep is expensive.