So, when using the various software to craft recipes, how should we calculate IBUs? Just look at the IBUs without the late extract addition? Or is the late extract addition calculation the one that is correct?
Uh, depends on the brew. My recommendations would be:
Use the full-boil, early-addition numbers as the closest to correct even if you're doing a partial boil. There may be some slight increased utilization from doing late-addition because the break material isn't all in there, but it's going to be a lot less difference than the formulas would say.
e.g. If you calculate Tinseth for addition 1 oz of hops at 60 minutes in a 1.050 full boil, and then try to get the same IBUs in a 2.5 gallon partial boil (1.100 OG) the formula would say that you need 1.6 oz of the same hops to reach the same IBU number. That's 60% more hops needed in the partial boil according to the formula. Home brewers have intuitively known for years that that's wrong, and have recommended rules of thumb like "10% more hops in the partial boil"--much more conservative corrections.
I'd ignore late additions and partial boils until you get a feel for things--certainly don't go with the massive changes the formulas indicate, maybe 10% more in a partial, 10% less in a late-addition, and treat a late-addition partial the same as a regular full boil. That's not exact, but we don't have enough data points yet to be very precise.
AND if you're formulating an original recipe:
Lie about the gravity slightly--calculate the recipe with no simple sugars (corn sugar/cane sugar/candi syrup/etc) and with double the "high-break" grains (wheat/rye/etc), calculate the IBUs, and then correct the recipe. e.g. if you have a hefe that gets 25 gravity points from wheat and 25 from barley for a 1.050 OG, lie and say that it's a 1.075 for calculating IBUS (double the wheat). If you have a belgian that gets 60 points from barley and 20 points from sugar, lie and say that it's a 1.060 for calculating IBUs.
BUT, be aware that if you're using someone else's recipe they've probably formulated it by taste rather than the numbers, so trust a well-formed recipe over any formula and don't futz with them until you feel like you have a handle on things. And if you're new to recipe formulation, look at similar recipes to get a handle on the amounts of hops that people have used successfully.
Also, don't expect to be exactly on-the-nose with IBU calculations. If you look at the numbers from that thread (where it tests the predictions of all the formulas vs. actual measured IBUs), they're all off significantly in some cases.
Probably
the most important thing to do is pick one method and stick with it, so that you become aware how a certain number is going to taste. And be aware of the wheat/rye and simple sugar issues Palmer raises.