Definitely. But there's also been very little investment into remote work, both from a research and infrastructure perspective.

I just wonder, if they were to take some of this cash and invest it into proving out remote work, would it turn out to be a better long term investment?

In most of the cases that I've observed, companies enter into remote work literally by just letting their employees start working remotely. Maybe they get a nice contract with Zoom and hire a specialized lawyer or two. But what if we were to really try investing in it, in the same way (or even in a fraction of the way) that we're investing in real estate?

Another point that I’m not sure gets much attention is that, as these companies build huge campuses and spread out to multiple locations across a region, how co-located are employees anyway? Sure you can move people around to be with primary teams but it’s certainly not like everyone can pop over to talk with someone or can depend on a lot of serendipitous hallway conversations.

That's a great point. Google tends to collocate teams by product, so that may be their answer to this. But it's true that you'll eventually reach a scale where you're going to have daily remote interactions, regardless of whether those people are in an office or not. Even at the Googleplex, if you're in an office across campus, it's better to connect to the meeting remotely than walk over.