But building your own doesn't scale to all the things. For everybody who wants to build their own X, the same person doesn't also want to build their Y and Z.
They will eventually need to buy some products. So there will generally always be a market for pre-packaged solutions.
For example: someone building an app may need network storage. They may not also want to block the building of the app on the building of the network storage.
If you want to build your own Dropbox with rsync, go wild, have fun, we'd all love to see what you come up with. But I don't have time for that. My family doesn't have the skills for that. Dropbox is great for us, and building our own is not a realistic alternative.
1) there exist viable commercial competitors providing approximately equivalent functionality
2) the roll your own solution with, e.g., TrueNAS, also provides equivalent functionality and is about 90% as easy.
I say this as someone who owns and manages three Synology boxes and one more recent TrueNAS box. There was a time when Synology offered something quite better than the alternatives, but that time is no longer.
My newest one (192TB) I bought the hardware pre-assembled and tested from a VAR, installed TrueNAS, and was off to the races. It cost more than buying the individual components would have, but it had zero headache and was cheaper than buying the equivalent amount of storage from Synology.
It's literally the "Why would you buy Dropbox when I can glue it together with rsync" level of ignorant comment, completely ignoring how behind most of those solutions like TrueNAS are in time cost.