I can't give you a definitive answer, but I can tell you this subject is rather complex and we can't state that retrofocus means (barrel) distorsion, while non-retrofocus means ruler-straight flagpoles at the edge of a one square meter print.
If you take a look at the diagram of the 3,5/75 Rolleiflex Planar, it looks rather similar to the 3,5/100 Hasselblad one and with the latter distorsion is slightly less than -0,1 % at infinity, however, the different focal length will change the distorsion figure, ceteris paribus. Take a look at any of the Schneider datasheets that show this clearly. Distorsion figures also depend on magnification factor (at both taking and enlarging). My Jupiter-3 1,5/50, a Sonnar copy is not retrofocal, yet it shows noticable pincushion distorsion up close.
If you take a look at the ZM 2/50 Planar, the 2,5/50 Summarit and 2/50 Summicron, not the same designs, but closely related, the Zeiss has roughly -1.5 %, the Summarit 1 %, while the Summicron aroung 0,25 % distorsion at infinity. While the 2,8/28 Elmarit ASPH is retrofocal, it has somewhat lower distorsion, than the C Biogon 2,8/35.
I would expect the higher end Rolleiflex lenses to have lower distorsion figures than my Xenar (because of more symmetrical design). Straight lines don't always stay so with my Rolleicord (with Xenar), but this is noticable much more with a ruler than to my naked eye, even on large prints. I was seriously worried about it for a week one time, than got over it. If you want low distorsion, cheap, I would suggest you find a camera with the two-element Rodenstock Periscope, but treat f/16 as its working aperture.
If money is no object, try to find a 60 mm Biogon equiped camera, until than the best choice would be to try a Rolleiflex and see if it's good enough for you.