That study is unfortunately obsolete since one or two weeks

Per the authors, it seems that that blockage in this study forms a “underpressure” on the chest, and creates weaker “overpressure” on the upper legs.

They don’t seem to get into hypothesizing how CdA might be improved over their best CF4 case, but based on their description, the stagnation is largely vertical and the “underpressure” would improve drag further if the torso were more upright. What I’m saying here is not necessarily that overall CdA would be better by raising the torso, but for cyclists that ride with less aggressive positions, the effect may be more significant.

Still, any improvement based on adding more losses (here via essentially separation off of the front of the chest) implies that there exists a better solution which mitigates that high velocity flow path from needing to be fixed at all… maybe by visors or hand and forearm placement.

I don’t really buy into the author’s justification for using RANS (steady simulations) as being sufficient, even though the author addresses not using SAS or DDES (transient, higher fidelity physics simulations). I also think the meshes near the rider+bike need a good deal more resolution to capture wakes better, even though they have near wall refinements.