I had proposed a starting point for a basic framework here:
I was basing it more on % of veBAL voting. We could also do ROI, but that fluctuates a lot more, is harder to calculate as a fixed number, and in the end a whale could just add more deposits and still capture more veBAL.
For example, how would you set this 100% ROI rule on the DIGG pool? Including AURA or not? Max boost or min boost? Take using what measurements? At what time? Remember that the ROI of a small pool can quickly be lowered by just depositing more assets (which whales can do).
I would rather discuss general frameworks than more individual requests to kill gauges though, so maybe we can resume a conversation over here: https://forum.balancer.fi/t/rfc-simple-framework-for-vebal-gauge-voting. Get something implemented, and then apply it consistently? If the DIGG gauge gets killed with nothing else done, we just move onto the next one and spend all our time doing this over and over again.