in general requires dissection and/or rearing. P. longipalpa (misspelled on BOLD) can be eliminated because it is only known from Florida, mining in a tree related to avocado.

According to Forbes (1923), which is cited as the original description for magnoliella even though the latter name was first used by Chambers decades earlier, liriodendronella and magnoliella are the only two species in NY with a longitudinal golden streak from base to middle. In liriodendronella the streak is partly edged with black; in magnoliella it is not at all edged with black. So based on that your ID is correct.

Incidentally, the only indication I have seen that these two species are not 100% faithful to their respective host plants is that Darlington (1945, Ent. News, p. 92), under the heading "Phyllocnistis liriodendronella Clemens and Phyllocnistis magnoliella Chambers," says: I have specimens reared on Liriodendron tulipifera and on Magnolia virginiana and cannot distinguish between them." -- but he doesn't describe the adults at all, so there is no way to know what he was looking at. I've seen Phyllocnistis mines in a magnolia bush 40 feet from my tuliptree, but haven't had a chance to rear them yet.

Davis & Wagner (2011), p. 44, lists three specimens of longipalpa. They have nearly identical barcodes and there are many specimens from as far north as Maryland with identical or nearly identical barcodes. Davis & Wagner compare barcodes to other Persea feeders but oddly do not mention magnoliella or liriodendronella as far as I can tell. The range of Persea borbonia, the host of longipalpa, extends from Florida west to Texas and north to Delaware according to the Wikipedia range map. But I'm not sure that's relevant because the specimens are mostly near the Appalachians and specimens from Ontario appear to barcode very close.

It would be reasonable for liriodendronella to be closely related to the Persea feeders since the host plants are somewhat related. The weevil Odontopus calceatus mines leaves of both Magnoliaceae and Lauraceae, though it isn't recorded from Persea specifically. I'd love to have my Phyllocnistis specimens from Liriodendron and Magnolia barcoded if somebody was willing to do it; maybe that would help clarify the situation. I've always been perplexed why Davis & Wagner attributed leaf mines found in South Carolina to P. subpersea when they gave no indication of how mines of that species would be distinguished from mines of P. longipalpa, which they said only occurs in Florida.

That was very helpful!

You might ask Terry and Bob to take another look... here is the original description of P. liriodendronella, and here is a mine like the one I reared this moth from. This is the only similar moth I see in Lyonetiidae (the species you had initially suggested), and it's only a superficial resemblance:

Maybe there's some difference between our two moths that I'm not seeing, but I'm certain mine is a Phyllocnistis.

Out of curiosity, are there any tulip trees planted near your house?