I appreciate your input, it's to the point and expands on the brainstorming I was looking for.
It's an interesting thought experiment, and it would be interesting to crunch the numbers and see what changes it would have on the DT1 list.
My thoughts exactly, emphasis on thought experiment and what the potential outcome would be.
If it were to be implemented, I would suggest changing it slightly. Let's say we have a user with 10 DT1 inclusions. As per your system, they could be excluded by 10 DT1 users and 20 DT2 users and 30 DT3 users and so on, and still be part of DT1. Surely it would make more sense to have an overall "inclusion/exclusion score". Each inclusion from a fellow DT1 would give +1 to that score. A DT1 exclusion would give -1, a DT2 exclusion would give -0.5, a DT3 exclusion would give -0.33, and so forth. They would need an overall score >0 to remain on DT1.
I like this for the simplification, it's very similar to the second part of my suggestion, but without the requirement for a minimum un-trusted threshold from one particular DT group.
It would also be much easier to calculate overall +/- trusted/untrusted values from DT members than my suggestions, from a math/programming perspective I imagine.
Having said all that, it would currently be open to massive abuse, and that would need addressed first. A DT1 user could simply add 100 sock puppets to their own trust list therefore making all those sock puppets DT2, and then exclude anyone they wanted from DT1. Perhaps some kind of counter voting system using the same or stricter merit requirements for the current voting system would work better?
Yes, I couldn't agree more. While sock puppets is a problem of sock puppets, it would easily destroy any attempts of DT decentralized accountability.
I also completely agree that a minimum merit/rank requirement would be necessary before considering to implement any of these suggestions.
This could still do with more brainstorming though, ideally from a more qualified mathematician in my mind.