This was a great description of what reality is for many startups, rather than some dogma from a book. I really admire what you’ve written here and what you and your team have achieved.
There are cases where you can’t validate what you’ve built until a substantial amount of technology is there. The minimum viable product is something that has to be engineered and implemented.
I also liked the fact that mParticle built to scale from the start. This is what I’ve done with nderground as well. I can’t afford to throw away the software and build a new version if it fails to scale, so I built it to scale from the start. If I can’t support millions of users, then there’s no point, I’ve failed.