Hit me up: What concerns do you have and how can I convince you the data is good? I'm obviously biased but I'm personally very surprised by how well it works and how much the sketching/metadata from exabytes of real backups contributes.

We have a lot of interesting results from research partners but most of it is unfortunately still under embargo. If you're in SF you're welcome to say hi (hit me up, my email is user at gmail .com )