I agree that just because people are looking down on PHP in general does not mean BGA should rewrite their entire framework to move away from PHP. Especially as there are long running projects where the developers do not care enough to migrate from PHP to another framework (because it means a lot of the work has to be redone). Additionally there is no reason that language popularity should decide what BGA uses for their own development, they should decide for themselves what brings them the best combination of development speed (including time spent for any conversion), quality and staff to work on it.
A potential way to enable support for the game implementations to be written in other languages is to decouple the game implementation from the rest of the framework by some IPC protocol (sockets, messaging queueing). This would require significant effort though so it's unlikely to be implemented any time soon. Additionally it makes it harder for BGA to review the code and provide their current framework features (like their own RNG implementation, deck handling and such). So even if BGA would decide to go this route sometime, it'd take them a while to set this up in such a way that they have the appropriate confidence that they can keep developing BGA in such a way that games don't break automatically.
