"Ideally they vote with their wallets and just leave Google, but again that's just easier said than done in many cases."
That's the ONLY thing that will ever make this change, but it rarely happens because "sunk cost" rationalizations (and "maybe it'll get better next year!").
Shouting from the rooftops for years has not worked. This is not anywhere near a 'new' problem.
Yes, possibly some people don't know it, and yeah, I totally get that "you'll just have to live with poor decisions made on your behalf.".
I live with poor decisions all the time (mine and others'). But the notion that we can bitch about it on forums, or make 'formal' requests for better support - none of that makes a lick of difference to them (or, demonstrably hasn't for several years). It's "google", so some people will flock to it because of the brand, or because it's not azure, or aws, or whatever. Or to tick the 'cloud' box on their spreadsheet. Or because google gave them thousands in free credits to switch.
"Some semblance of professionalism shouldn't be too much to ask from one of the world's biggest corporations, an ask coming from paying customers no less."
But... it demonstrably is, and has been.
And... you'll get a client who will say "I can't believe this, email someone, get this fixed". At some point you have to push back and say "no, this is simply not fixable, you have to live with XYZ, or pay the cost of moving somewhere else".
An extremely large customer, maybe, possibly, might have a bit of say, perhaps. But even if you're dropping millions with them... for a company that measures revenue in hundreds of millions of dollars per day, you probably won't hold much sway. If you had that sort of budget to spend, and it was that critical, you'd likely be doing enough due diligence that research in to actual support/service levels (vs what the sales contact tells you) would be on the agenda.
Vote with the wallet is the only viable/impactful approach, and people have to be willing to forgo the sunk costs.
And for people that don't know up front - they always had the opportunity to search (google) beforehand. If they choose not to - caveat emptor.
10 years ago, we could possibly be forgiven for choosing google's services (GAE, etc) because there wasn't as much of a track record to look at re: support, etc. There's no good excuse for starting a project today on google compute and claiming ignorance of their support levels.