I also would recommend k-omega because it is the only system which has a clean phyiscs basis, see attachment (published in Physica Scripta 3 years ago). With the SST add-on I personally have no experiences, but a counter questions: what is your Reynolds number based on mean-flow data and diameter of the "pipe"?
Article Universal equations and constants of turbulent motion
I have transition stage where Re number varies. It is the case of flow over transonic rotor blade. I will go through the paper attached by you and will let you know if i have any further doubts.