Regarding factoring the resists respectively to their max, instead of doing an average of the resists, as you will give the max resists of each pets and their minimum possible resists. The 65% of the resist count could represent the average position of each resist in regard of it's highest value.
I sent Nexus how I do the calculation for the pet power calculator, and it is basically what you described. For each type of pet, I determine the range of max and min possible resists. Then, the pet's own resists are compared to that range. That way, the range is specific to each creature - because like you said, it wouldn't be appropriate to compare a resist value from a pet with a very small range to a large range of resists.
With weighting, I've got an experimental weighting feature only applied to greater dragons right now on the pet calc (I described it above in a previous post in case Nexus would like to do the same) - instead of straight averaging resists, a pet's highest resists are weighted more heavily in the 65% overall resist weight than the pet's lowest resists, using the thinking that you wouldn't likely choose a pet to fight a monster with high damage in the pet's lowest resists, making those resists less valuable overall. I'd like to extend that idea to skills, at least for pets that have distinguishing skills (like poisoning on a rune beetle).
There's a thread in the Tamer section here for feedback of the new weighting calculation, if you're interested to try it out!