For most of the 20th century the name "V. palmata" was thought to apply to plants consistently producing lobed leaves throughout the growing season. "V. triloba" was used for plants that produced unlobed leaves in early spring (and sometimes also in autumn).
In the 1990s, several botanists independently realized that the names "V. palmata" and "V. triloba" had been misapplied: it turned out that Linnaeus's "V. palmata" actually had some unlobed leaves, so this name should have been used to refer to such plants. "V. subsinuata" was chosen as the replacement name for most of the plants with always-lobed leaves that "V. palmata" had been referring to.
Most iNat users seem to be using these names in the post-1990s sense, anyway. My suspicion is that much of the continued use of "V. triloba" is meant to indicate the shallow-lobed variety, V. palmata var. triloba. But there are also plenty of observations with deeply-cut leaves as well.
Also, some observations of plants with all leaves deeply-cut currently have iNat identifications of "V. palmata". These IDs might reflect the older use of the name, but they could also indicate a preference for a broader species concept.
See the flag linked-to above for more about the varieties, some information about active research on these plants, and the rationale for including the varieties despite Kew's POWO not having pages for them.
Ballard, Harvey E., Kartesz, John T., Nishino, Misako. A taxonomic treatment of the violets (Violaceae) of the northeastern United States and adjacent Canada. 2023. J. Torr. Bot. Soc. 150(1): 3-266. doi.org/10.3159/TORREY-D-22-00029.1 (Vínculo)
Los desacuerdos no intencionados ocurren cuando un grupo padre (B) se adelgaza al cambiar un grupo hijo (E) a otra parte del árbol taxonómico, provocando que las Identificaciones existentes del grupo padre sean interpretados como desacuerdos con las Identificaciones existentes del grupo hijo cambiado.
Identification
La ID 2 del taxón E será un desacuerdo no intencionado con la ID 1 del taxón B después del intercambio de ancestros
Si el adelgazamiento del grupo padre provoca más de 10 desacuerdos no intencionados, deberías dividir el grupo padre después de intercambiar el grupo hijo para substituir las identificaciones existentes del grupo padre (B) con identificaciones con las que no esté en desacuerdo,