Riegel's exponent of 1.06 was fitted to records from 200 m to the marathon. It works well over one or two steps of distance. Predicting a marathon from a 5 km time stretches it a long way, and the usual failure is that it is too optimistic — endurance does not scale as neatly as speed. Treat anything more than about four times your input distance as an upper bound on your ability, not a target.
Training paces are derived from the predictions, roughly on Daniels' model: easy running at about 1.3 times your 10 km pace, threshold at half-marathon to 15 km pace, intervals at 5 km pace, repetitions faster still. The single biggest mistake in amateur training is running the easy days too hard, which is exactly why the easy pace here will feel embarrassingly slow.