The equations I've looked at so far don't really account for massive late hopping and whirlpool additions. I really like the flavor and aroma I get from late hopping, but It's been very hard to brew a consistent beer. I realize that none of these equations is very accurate to begin with, but from a recipe design standpoint I really need a way to estimate how much bitterness I'll end up with. How do I (somewhat) estimate the bitterness impact of super late additions?
For whirlpool hops, you don't have to worry about increasing the IBUs since the alpha acids are barely soluble in their natural state, and contribute bitterness only by being continually boiled.
For late additions, since these are boiled, then the usual IBU formulae can be used. John Palmer gives Tinseth's formula and computed values for a variety of times and SGs. There's also a discussion of the formula from 'Designing Great Beers' by Ray Daniels here on HB stackexchange. For additions less than 10 minutes, 0.06 utilization is assumed. (The link gives the formula so you can compensate also for gravity if above 1.050.)
But as the OP mentions, none of the formulae are accurate. They are best used as a rough guide, which you then refine by inspection of the final result, and adjusting the amount of late additions accordingly. One thing to bear in mind is that it's believed that utilization decreases as more hop mass is added, but the current formulae don't take this into account.