I would categorize data tooling as a moving target - we have some, it's never enough, it probably won't ever be enough. It's a moving target (p.s. obligatory we're hiring!).
More generally speaking, (excuse me for being a little hand-wavey here) many of the AWS offerings feel like polished, managed versions of familiar tools. Redshift, for example, feels a bit like "hey we figured out how to abstract away a bunch of mysql instances to feel like a big processing cluster". That's not a bad thing, necessarily. The google stuff feels much more intentional - "we need to solve the problem of doing these sorts of queries at scale" vs. "we need to solve the problem of scaling mysql to solve these types of queries"
Maybe they're just better at abstraction, but whatever - that works for me!
Nice! Did you have the data tooling built out before you went to Google Cloud? If you did, I could imagine the migration was pretty hard as well.
Also, all of those seem relatively possible with AWS Redshift, Kinesis, and Data Pipelines. I'm interested what Google Cloud had to offer, spec-wise.