Don’t think theres a real conclusion, that can be made by reasoning it out.
You’re opinion here will hinge on weather you believe the BGE is generally useful and the perceived impact of the trade-off we make by keeping & maintaining it (which we can’t really quantify - but its not zero).
If you use the BGE and ran into some bugs/limitations which prevented you from making a game… that will influence opinions too.
Also if you tried to work on improving the BGE and ran into really bad-code or bugs which were hard to fix that seem like they should have been easy… it will influence your opinion.
We’ve been discussing this (in #blendercoders IRC)
The likely outcome here is - Blender game engine stays in maintenance-mode (just keep it running). And some developers will investigating supporting an external engine (Using similar API hooks we use for external render engines), then if this integration can be made to work about as well as the BGE, we can deprecate and eventually remove the BGE. But this is a way off, and just something to investigate, no firm decisions made yet. Its also possible someone works towards big improvements, But so far this didn’t happen much.