Speaking in broad generalities, it's generally more conducive to precision for the bullets to be AT or slightly over groove diameter from what I've seen. That being said, I've seen plenty of .3078" bullets that shot amazingly. They still obturate and swell to the bore under the pressure load from expanding gas behind them so that's far from a hard and fast rule. I've shot .308 bullets in 7.62x54r before and while the rifle wasn't capable of incredible precision by modern standards, those .308 bullets shot just as good as .312" bullets, probably near the limit of what the system was capable of.
In general, I avoid tight bores because they do cause more heat and more jacket disruption, which can increase risk of bullet failures-- again that's a broad generality though. Lots of them are just fine with no issues even with super thin jackets. All of this is bell curve distributions and overlap and only the tails of the distribution crosses over thresholds so it's percentages of percentages-- hard to speak definitively. I would personally have less issues with tight groove diameter vs. tight land/bore diameter, but that's not based on anything scientific, just a hunch.
All else being equal (rifling sharpness, profile, count, ratio, etc...) half a thou on groove and 2 thou on the lands really shouldn't matter much at all unless you're at the far ends of spin rate, jacket thickness, pressure, velocity, heat, etc.