It seems to me that where JetBrains went wrong was building every language supported as a separate application. According to their announcement post the stated goal of was to deal with the customer ask to get easier discounts on multiple language products or to move from one language to another more easily. To JetBrains the solution was apparently to switch to SaaS, which kind of makes sense, but from my perspective maybe they should have just consolidated some of their product efforts?
The subdivision served multiple purposes: it increased revenue by segmenting the market, and it allowed for a level of specialization that was difficult to achieve with a single tool. For example, some of the PyCharm stuff requires hacks that most Java developers wouldn't want in their IDE, and I honestly wouldn't want to deal with any Java stuff if I can avoid it.
At the end of the day, the announcement is just corporate speak for "we need more money more regularly, so from now on you're going to pay rent; to make it a bit sweeter, rent includes use of swimming pool and sauna, which you may or may not need". Most businesses switching to SaaS don't do it for the customer, they do it for themselves.
It does through plugins. The language-specific IDEs are updated with language features earlier than the plugins for IDEA and they have simplified project creation/management. For instance, it's a lot easier to setup a Python project and choose your interpreter/packages in PyCharm than it is in IDEA.
