Mainly because we wanted to start off a small number of annotations which were applicable to large numbers of commonly observed taxa, so we went with a few broad ones and got caught up in working on other things (mainly dealing with the continual growth of iNat).
This has been discussed elsewhere on the forum. TL; DR: if it’s going to happen, it’s not going to happen anytime soon, although it’s a cool idea, and the more annotated observations we have to train a hypothetical model on, the better, so it would be great if people can continue annotating observations. You can use the Identify page to do that pretty quickly, I made a tutorial here.