It would be nice to have a metric which maps validator fees and realised availability to a single metric, which can be compared against all active validators. Given the all or nothing rewards aspect for delegators, it would be good to see a metric which allows people to see a clearer distinction between headline fee and availability (i.e. a low fee validator with low availability may be worse than an expensive validator). For example:
For a given stake of K Near, compute the number of epochs required to earn another K Near (2x stake), net of fees, not accounting for compounding, and assuming 100% availability for each validator
Now add on extra q(i) epochs required to epoch values computed for each validator in 1) that takes into account the number of expected to be missed based on historical realised availability such that the net staking reward (net of fees, availability adjustment) is at least 2K Near.