I do believe an additional metric that included frequency of purchasing botvotes would better
This isn't a bad idea.
You're right, I basically agree with the concept on the whole, but I'm simply taking into account the reality of the likelihood of it being effective. We know from haejin that almost nobody is willing to sustain high-value flags to accounts, let alone getting $500 down to 0. There's just nobody on the platform with that kinda money and that kinda focus, and those who do care take thousands of flags to get rid of a couple of bucks.
This is not going to change any time soon and no initiative of self-policing is going to consistently cause any impact whatsoever - consider 'flagawhale', probably the largest community of flaggers for this cause that are, together, imperceptibly ineffective for the most part. Haejin still makes as much as he ever made (minus steem value) and bots users and owners will continue to profit.
The problem of bots needs to be disincentivized from the core, not from self policing.
With that in mind, blind flagging is not an effective methodology in my opinion, because nobody will want to waste their precious SP, whereas flagging based on your subjective consideration of value upon viewing a post, is something that should be an active part of the steemit culture. The rest is an inherent flaw that needs fixing by STINC somehow.
The thing is, even if we get all the current bots to agree on certain rules and regulations, it's a verbal contract and thus hardly binding, and any new bot services that prop up are going to instantly see how to take advantage, and do so. All it takes is a single millionaire to come on board and we have ourselves yet another crisis we spend weeks on battling with futility until we forget about it, until another millionaire comes on and repeat.