I think moths must be the most problematic group I've encountered.
First of all, there are over 2000 species to consider. Then, as I've mentioned before, colour is generally much less important than pattern. We also have to factor in the fact that some species are extremely variable. Then, finally, we have to consider wear.
If you have an unidentified specimen in front of you and a reference book (or more than one book!), then just trawling through the images won't help very much. It's a matter of finding something that:
a) looks quite similar
b) has a phenology that more or less matches.
c) is a reasonable size.
d) has plant dependencies that match the location where the specimen is found.
Then you need to read the species accounts for the candidates, and see which of the critical separating characteristics are present. References to 'similar species' should also be consulted.
That should lead to an identification in perhaps 90-95% of specimens. The remainder might be too worn to sustain an identification, but we also need to then consider aberrations.
The following specimen has been identified (not by me!) to be Dotted Clay - Xestia baja.
The species account says: "The most conspicuous diagnostic markings are the two small, sharp black dots just below the leading edge, close to tip of forewing." Sadly, mine has only one "sharp black dot", and one blurred one. So when all else fails, we need to invoke Dunlops "fourth law of moth identification". Which goes something like this: "When a moth fails to satisfy the basic diagnostic markings of all species, we need to add in missing features one at a time to see if they lead us to an identification, as long as they don't conflict with the diagnostic markings of any other species." Which means that in order to identify all specimens satisfactorily, we need to know what all the rest look like, or might look like.
This might take some time.