We know that at the center of every significant galaxy, lies a supermassive blackhole spinning at extreme speeds. This spin also drags space with an angular speed inversely proportional to the cube of the distance, converging to the speed of light near the event horizon. In the flow model, this introduces a tangential component for the gravitational acceleration vector, which gets comparable and even surpasses the radial component, as one gets closer to the blackhole. (An anology can be made with whirling water in a sink). This results in a smaller than expected (radial) gravitational effect and thus less orbital speed for nearby stars, as opposed to an (almost) pure radial vector and normal orbital speeds, for stars at the outskirts of the galaxy. Relevant equations can be derived by adapting the flow vector to Kerr Metric. This approach has important consequences:
1) Central blackholes (galactic centers) are much heavier than previously calculated.
2) Stars at the outskirts (right end of the galaxy rotation curve) give a better indication of galactic center’s mass.
3) Dark matter is not necessary to explain flat galaxy rotation curves and extra gravitational lensing.
Same reasoning is also valid for galaxy clusters. Please note that the above summarized approach can also be applied to General Relativity via skewed/warped gravitational field lines due to frame dragging, albeit with more limited effects. It is also highly possible that frame dragging has a significant effect on the shape of (especially) spiral galaxies.
This model employed some reverse engineering and math is not presented at this stage.