Hey guys,
This thread is shifting to many interesting insights. However, this topic has a clear proposition of killing a gauge, which should be just as simple as enabling one. It’s up to veBAL to decide which gauges are “playing the game”.
My experience in Balancer so far, has shown that DeFi DAOs have a much bigger need of constant iterations, moving quickly and staying vigilant. While other social and even investment DAOs can give way more time for experiments to flourish.
I see @Andrea81 point, and maybe we would be in a different position if we had ~20 projects like Cream diluting voting power between pools. It would be nice if they were generating fees, tho. 
To me, the case is built around what @SmallCapScience brought up here as “we are looking at this the wrong way by highlighting income from trading fees”, and we should consider what role Balancer is playing in this. Personally, I think there are plenty of motives for us to kill the gauge, atm (not final, as @zekraken pointed out).