Did you actually see a quote from Kx before making that decision?

I've first-hand seen the cost/proprietary argument thrown into the face of KDB, to then see companies proceed down a path of other technologies, with a 10x spend in hardware and dev effort.

KDB is insanely tuned.

There are other costs. It is hard enough to find people in my industry with domain knowledge, ability to dig into problems, and SQL skills. Finding enough people to use an array language would be difficult as others in my industry have found. I'd personally be super pumped, but I'm an outlier here. Also, there are costs to getting kicked into something like this instead of something like tsdb or influx.