Realistically, there's no way a typical end-user is going to deploy their own seed at all. I suspect they know this, as suggested by the fact that they plan to

Realistically, there's no way a typical end-user is going to deploy their own seed at all. I suspect they know this, as suggested by the fact that they plan to build a hosting service. (I presume that's how they're planning to monetize this, long term?) Anything that requires active system administration or application deployment by the end user is going to fail. IMHO, I'd consider language/framework choice to be the least of their concerns; they picked something they knew, and moved on to more pressing issues.