"Tell me what would be necessary, within reason, to make a man so visually acute that they see in a framework of nanoseconds and react accordingly fast. Like the ultimate gunslinger if it were an old west type like Wild bill Hickcock or Doc Holiday, Wyatt Earp ect. Based on what I've given."
The main thing would be to make the neural conduction velocity be light speed rather than sound-speed (3e8m/s rather than ~30m/s), in both the CNS and peripheral nerves. The synaptic delays would be the next thing to handle. I'm not sure how well it would work to have "legacy" parts of the brain operating much slower. It might not work at all. Even with near-instantaneous reactions, movement is bound by physics. On the plus side, humans use their muscles much more gently than other animals, including hominids, so there is some headroom for additional force to make fast movements. Additional efferent muscle innervation might help, with blocks on excessive afferent pain signals that usually prevent high muscle forces in humans (this sort of temporary block is likely responsible for hysterical strength). A predominance of fast-twitch muscle and anaerobic metabolism with the other modifications could allow up to 10x faster movements for periods of a few seconds per minute. To make precision movements would require dialing down the efferent nerve signals, though.
Human eyes are optically lousy. With good optics they can get down to 20/8, for more, the size of the rods and cones would have to be reduced - but with better processing and slight movements of the eye, perhaps 20/4 or better could be possible. With massively more sensors, making the rod/cone density equal to the fovea across the whole field of vision, rather than just 0.5 degree, the amount of visual data would be staggering, several thousand times what an ordinary human sees. It would theoretically be possible for each photoreceptor to respond to light from say 200nm (shorter waves would be difficult- not much is transparent at those wavelengths) to 1600 or 3200nm, (resolution falls off with longer waves, need bigger photoreceptors), so that's 3 or 4 octaves of light. (3 octaves from 300 to 2400nm would be a fairly conservative range) The receptors could also theoretically tell the wavelength and polarization of every photon, allowing the person to see the specrograph of everything in his field of vision, and thus its material and often its temperature, at least for very hot things. (He'd likely see the tail of the blackbody distribution at surprisingly low temperatures, given his spectral resolution, but the peak of the distribution would only come into view at several hundred degrees F). A slight increase in the size of the eye (up to about 1/3 bigger) wouldn't look too weird and would increase spatial resolution. Being able to distort the lenses more, and more precisely, being able to distort the shape of the eye itself, and perhaps having additional internal flexible, movable optics inside the eye would allow a zoom lens effect up to perhaps 2 or 3x. Having no iris and an exceptionally wide aperture would give a very thin focal region, and thus potentially a very precise perception of distance. Being able to refocus rapidly and doing so constantly and unconsciously would give an extraordinary 3-D situational awareness. When "going into overdrive" and effectively seeing thousands of "frames" per second, in ordinary indoor light, things will seem to go dim and grainy for him, as they do for us in very low light at ordinary perceptual speeds. On the other hand, a good deal of color vision should persist in any light.
So to sum up, with the eye completely redesigned, perhaps 20/1 vision in the near UV to 20/8 vision in the mid-IR, covering the whole field of view (about 80000 times the area of the foveal solid angle), with the ability to see thousands of separate, independent colors and polarization, a 2 or 3x zoom, absolute distance perception accurate to about a millimeter at 1 meter to a few centimeters at 100m, on the order of 1e7times more information per "frame" and thus a trade-off in using processing power - even with a 1e7 faster visual cortex, ramping up to a faster "frame rate" will degrade the fineness of the processing and the amount of data that can be handled (thus relative tunnel vision). Also at high speeds, there just usually isn't enough light for smooth, bright vision, so he will have dimmer, lower-contrast, grainier vision at high speeds.
You may find http://www.nanomedicine.com/NMI.htm" by Robert Freitas a useful reference. An additional volume is available at the same site.
Also see the free PDF http://www.nap.edu/catalog.php?record_id=12896" from the National Academies Press / Defense Intelligence Agency
