[mirimir]

That's a different sort of "t", isn't it? As in the lifetime of a state, rather than a measurement of duration.

[mablap]

It is indeed, because "time" is not an operator in QM, unlike H or x or p. But from my limited understanding of the time-energy uncertainty relation, you could (as you say) consider it a measure of lifetime - the time it takes for a change in E . So if you get a higher energy, you also should get less change in time, and thus you could probe smaller timeframes.

I might be wrong though. Someone shoot that argument down if you have to ;)

[M_Grey]

The spacetime interval should be invariant across frames, so there should be a problem there.