I think the problem this product solves falls into the chasm between teams that can fix this in a few hours themselves and people who don’t know what DNS is. Unfortunately there’s not many people in that gap.
Looking back, I shouldn't have ever scaled it beyond my personal needs, and kept the architecture simpler, and not "highly available". Cost would have been only 10% since that service could run on a single $5 droplet running Caddy server