It's just not the right setup for a post editor. It's okay as a site builder if you put a decent amount of work into creating your own blocks and what not, and I suspect it may have a niche on very large media sites which have a bunch of overly complicated article layouts.
But for your average Joe, it's just awkward and makes posting simple posts more tedious/waste more time than it needs to be.
It also shouldn't have been merged into core already either, given the software clearly isn't ready and basic things like accessibility aren't up to par. It's also really poor timing given that many sites are getting ready for Christmas, and retailers likely won't want to have to teach their staff a whole new editor in order to add products or deals at the most lucrative time of year.
God help any agencies too. Those guys will be working round the clock to try and fix client sites broken by the new editor (or angry customers yelling at them about how they can't figure out how to write content any more).
Not a good move guys.
It's also way too 'big' in general; the UI is a pain to scroll past and find things in. Definitely gonna go back to the Classic Editor if the next few posts don't go better than that one.
But it's also been fun to try and come up with some tooling so we can automatically install the classic editor plugin on the 300 or so sites that we manage on a variety of hosts, with a variety of methods of managing them.
What's been extra fun is that one of our main hosts is finally forcing everyone from PHP5.6 to PHP7.2 this month, so I've been having to go in and hand test every site and install shims for stuff like mysql_real_escape_string.
To be clear, I think that both GB and PHP7.2 are necessary advancements. I just wish that I had better testing strategies in place... but that is my own fault. I've been lighting candles for the folks who have fewer technical skills and resources than my small company has.
The new block editor doesn't offer anything for our purposes, and in fact breaks most existing sites built as content management systems (custom post types and fields). Sure, they say I can install their "classic editor" plugin on all the sites - but, aside from sites of clients with whom I've lost touch (in which case the sites will just break upon core update), I'm disappointed in how they've rolled out this change seemingly without regard to a significant portion of the community raising concerns. I mean, just look at the reviews in Gutenberg's plugin repo..
For me, it means using a fork of the last "good" version (4.9.8) - possibly ClassicPress - for the near term, and transitioning to other solutions (static sites, React/Preact fullstack) over the long term. During the process, we plan to "encapsulate" WordPress for use as a headless CMS, as a data backend to be swapped out eventually.