I'm not going to go into all the reasons why a multi-cloud solution probably isn't and why multi-cloud is stupendously difficult (and expensive as you move all that data around) for a scaling business.
But in a situation like this, a company that's likely to be deplatformed, is likely to be deplatformed everywhere. I've seen a screenshot recently of Parler's hardware requirements (when searching new vendors) and it's somewhere in the vicinity of 15,000 cores with more than 400Gbps (sustained) of internal and 100Gbps of external bandwidth (from memory).
The big cloud providers are the only place you can move to when you have those types of requirements (and are still actively growing).
That's about 250 bare metal 64-core servers. You can get that elsewhere, and keep in mind per-core performance on bare metal is going to be higher than you get per-core on shared tenant VMs in the cloud (for several reasons). Those bandwidth needs would constrain you only to larger scale bare metal providers though, which would leave you with fewer options.
The problem sounds like it would have been database sync between locations, which is a major issue in multi-cloud. Most replicated databases are pretty chatty.
You're right though about deplatforming, and the problem is if you are providing a safe haven for a bunch of Nazis like Parler is you are going to be constantly DDOSed. That means you need a lot of DDOS protection, and high end DDOS protection is a significantly smaller market with fewer players than just raw rack-em-and-stack-em hosting.
But in a situation like this, a company that's likely to be deplatformed, is likely to be deplatformed everywhere. I've seen a screenshot recently of Parler's hardware requirements (when searching new vendors) and it's somewhere in the vicinity of 15,000 cores with more than 400Gbps (sustained) of internal and 100Gbps of external bandwidth (from memory).
The big cloud providers are the only place you can move to when you have those types of requirements (and are still actively growing).