Services also tend to be better if lock-in effects on customers are low. ... and using your own domain for email does reduce lock-in significantly.
unrelated: the link in your profile does not seem to work.
But there is a case for making legislation forcing email providers to allow moving emails to other providers (how it should be done technically is another question). This is already in effect for telephone numbers many places.
As such, migrating these email addresses was easy enough.
My older @gmail and @googlemail ones though, not so easy. I've been moving each account I have used with these addresses one-by-one, but you never catch them all and even when you do, some services simply will not let you change your email address.
I recall being so excited when Gmail first launched and was one of the first people to get a Beta invite. I regret ever signing up for them now, given the headache it has been to get off it.
The main thing here is to avoid single point of failure as in both domains (politically induced problems) and infrastructure (technically induced problems). If people would use more than just a handful domains/providers for mail then single failures would not have that big of an impact.
I have been hosting my mail on my own domain for the past 3 years and have not been impacted by this incident. Currently I am happy for protonmail that I use to host my mails at. But I know that I can easily move on, if service drops, and even selfhost.
and the weight of google as well!
an individual mail getting onto an blacklist is most often than not a dead sentence for the address, the domain and sometimes even the ip.
but if google is at fault and email get into a permanently removed bucket, like in this event, it's in the interest of the other to play nice and accommodate for the fault.
I think people severely underestimate how hard it actually is to consistently deliver email in 2020 between dkim, spf and domain keys while tiptoeing around everyone else ip/email antispam services.