1) There is a correspondence, but there are more restrictions on YT than in the case of SU(N) where the restriction is not to have more than N-1 boxes in a column. Due to an additional invariant tensor, metric δab, we can take traces and irreducible tensors need to be traceless. Combining it with the ϵ-symbol you get that the total height of the first two columns has to be not greater than N, otherwise the tensor is identically zero.
2) There is a generalization of the Hook formula that works for SU(N). The details can be found in this old paper.
3) Again there is an algorithm, which is a superposition of two SU(N) tensor product rules. One has to take into account traces. The precise formulas are the very first in this paper