Unfortunately, being aware of the planning fallacy does not make it go away.

True. But there are ways to calibrate for it. It seems that subtracting off 10-15% for technological predictions works well. If one is being more careful it probably would do something that was more careful, say taking not a fixed percentage but something that became less severe as the probability estimate of the event went up, so that one could still have genuinely high confidence intervals. But if one is in doubt simply reducing the probability until it doesn't look like the planning fallacy is likely is one way to approach things.