I've been through some weld validations and there are so many variables that it has always been a case by case basis determined during the planning stages. But often what we thought we would look at during the early planning stages changed when we did the runoff as problems arose which caused us to change what we thought would be appropriate. Differences such as metal thickness, alloy, etc make it hard to give a blanket answer.
If I read you correctly, you're saying you change tips on all at the same time but some tips weld more spots (and/or at a higher current) than others. The assumption would be the tips with the most welds (or higher current) would guide your replacement PM schedule. I don't see that as part of the validation issue.
In similar situations I looked at several variables at the end of the process:
1. Correct number of welds per part and each side or piece.
2. Welds each tested (pull apart or pushout - depending on the part).
3. Visual or cross section (depending on criticality of weld) for weld penetration.
4. Within dimensional tolerence (sometimes an external fixture but I have seen the weld holding fixture its self calibrated and used).
5. All parts /holes present.
That would be validating the group of cells. Whether or not you would want to look at individual cells and their individual steps would be something to consider but it depends upon the situation. During trial runs I'd look closely at each step, but normally I would validate the process as a 'system' - or at least base inspection checks on the process as a system. I would watch internal and external nonconformance feedback and any that point to failures in the weld operations might lead me to want to investigate and revalidate each cell / step.
Putting dimensional aspects aside, the requirement is probably a weld separation issue such as xx pounds to pull the assembly apart. Many times assemblies are designed with enough over engineering that one weld of five failing to completely penetrate (as an example) will not cause the assembly to fail the pull apart test. Center welds are often less critical because end and side welds, if penetration is not complete, will tend to allow a part to bend if it 'unsticks', if the metal is somewhat thin.
I have seen issues where a welder was set for a higher current at one step than another using the same weld head. A problem was the current difference was enough to make it difficult to balance - one spot would burn and the other would be OK, OR one spot would be OK while the other often didn't penetrate.
These are just some thoughts.