Cultivars certainly are taxa. But they belong to a system of taxonomy which is defined completely separately from the systems for wild organisms (although it’s built upon and inspired by the ICN, the ICN in turn does not refer back to it). Given that inat is specifically about observing wild organisms, I don’t think it would be appropriate to incorporate a whole separate taxonomy system that is explicitly only for cultivated plants.
I can see that it might occasionally be interesting, like you said to observe which cultivars can end up escaping into the wild. But I think the amount of confusion it would cause - both in identifying individual observations and in users getting the impression than inat is intended for observing cultivated plants - would cause far more downsides than upsides.