Ah, I see. Perhaps I misunderstood what you were saying originally. So, to be clear, you're saying the current build of 4.8.0 doesn't actually introduce the fix I introduced here, but rather, introducing the fix I demonstrate here may become more complicated with 4.8.0. Right?
I'm hopeful the good folks at GL.iNet can implement the fix for everyone once they've had a chance to review my feedback (many of them are on holiday until the 5th of May from what I've heard). In their most recent correspondence to me, they said they'd likely respond within this thread (I made them aware I'd posted here as well). I did speak with someone who originally said the way it is right now is by design, but I think that's incorrect. Perhaps we were just crossing wires a bit in that communication / not communicating clearly with one another, but in the world of "features and bugs", there's no way anyone can convince me that an implementation that introduces DNS leaks is a feature. 
This sounds amazing! Extremely interested in this and grateful to see the GL.iNet team introducing this!
I agree, but what I'm asking for doesn't seem terribly complicated. I've started a new post about that if you're interested: GL.iNet UI shows no clients connected due to use of VLANs - #3 by Integritas.