Here's another piece of conventional wisdom of which I'm very skeptical.
Lots of people suggest fishing nymphs before a hatch, emergers early on, then low-floating dun imitations, and then high-floating dun imitations to imitate fluttering dry-winged duns. It's as if they're trying to imitate the emergence of a single insect spread out onto the timescale of an entire hatch.
Here's an example from
Mayflies by Knopp & Cormier, describing how trout react to a BWO hatch:
As the hatch continues and the concentration of emergers in the surface film increases, trout usually respond by slowly switching their feeding pattern from the floating nymphs to the freshly emerging duns.
It just doesn't make sense.
A single mayfly will take anywhere from one second to several minutes to pop out of the water, dry its wings, and fly away. Throughout the duration of the hatch there are many insects in every stage of emergence. If you think about the math, the ratio of emergers to low-floating-duns to high-floating duns on the water at any given time should not change very much over the course of the hatch. Even if they did change, the trout sampling random one-by-one encounters would not be able to notice and adjust to the change in ratios immediately.
I think it makes more sense to choose an imitation based on the duration of each stage of the emergence for the species you're imitating, regardless of whether it's early or late in the hatch (with the one exception that it makes perfect sense to fish the nymphs in the hours leading up to the emergence). That factor should influence what the trout are seeing
much more strongly than whether it's the beginning or the end of the daily hatch.
There probably are some trout feeding on emergers, some on duns, and some on nymphs throughout the hatch. But it's probably determined by the luck of the draw and the fish's personal experience with the more vulnerable stages of the insect. Fish are bound to see different frequencies of floating nymphs, emergers, and duns in different holding lies.