Percentile mitigation will be balanced around defensive (curative or absorptive) potency by selecting a particular 'standard' of DTPS (damage taken per second) likely to occur over a given span.
If we take a higher DTPS standard, such as Expert Roulette double-pulls, then potency hold up against percentile mitigation even in those situations while potency is superior if the DTPS is lower than that standard and percentiles are superior if the DTPS is higher.
So again, you pick a situation as your standard, such that percentile mitigation over 8 seconds of that situation should nullify as much damage as the potencies would absorb or heal.
Given that standard, would the 40 seconds per minute of 10% mitigation that the OP has suggested be traited atop TBN exceed the value of the 8-9.6 seconds per minute of mitigation and potencies of its competing traits? That's it.
I'd need a standard by which to find pre-active-mitigation DTPS and determine whether that the suggested buff would be overkill. Ideally, I'd like a rationale, too, as to why that standard would make sense (e.g., because it renders more similar mitigation+healing over time between jobs and therefore seems the more likely intended balancing point). But that's the whole of it. It's either noticeably greater than its competing trait's values and may be OP, noticeably lower and UP, or it's roughly balanced. But the best way to figure that out is simply to compare this suggested trait against the other traits. DRK's lack of self-sustain, etc., should be addressed separately as such, not as part of its lv82 trait.