Is it possible to create an artificial vision system with high-res
depth-of-field-capabilities, 10-100x on-thought zoom (the machine will
zoom as you think you want to zoom), blending several perceptual modes
(passive visual, EM, IR, UV, and why not, active X-ray) and spanning
the full 360 degrees something a man-machine system could send to the
visual centers and have a user experience consciously?
In order for photons to be detected, they must be absorbed. Furthermore, they must be capable of triggering a chemical reaction or electron cascade detectable by biological or semiconductor instrumentation.
One problem with long wavelength photons, is that they can only be absorbed by objects whose minimum dimensions approximate dimensions of that wavelength. Putting this another way, in order to detect $1 cm$ wavelength radio, you need an antenna approximately $1 cm$ in size.
This size restriction puts severe constraints on the resolution possible with many radio wavelengths. A human eye's dimensions might be capable of detecting 1 pixel of information for this wavelength and could not detect any radio waves longer than this.
It would seem then that we should be able to get very high resolution images from very short wavelength photons and, in theory, this is true. However, very short wavelength photons possess very high energy levels. This makes them difficult to focus (by reflection or refraction) and becomes nearly impossible at very short wavelengths.
What is worse, photons must be completely absorbed or they can't be absorbed at all. What this means is that electrons in the atoms of our eyes, simply can't absorb this much energy and the photon passes right by them. These very short frequency ($\gamma$ rays) react mostly or only with the nucleus of the atom. When the do so, they tend to deliver enough energy to change the composition or structure of the nucleus and break bits off.
In layman's terms, they cause the nucleus to fission. If the nucleus is more massive than iron (atomic mass of 56) the fission event releases even more energy.
Regardless of whether the $\gamma$ ray absorption is endothermic or exothermic, it causes your material to transmute all nearby elements into other elements.
This could be a problem for you when it happens in your skull.
Some of what you propose is difficult or impossible to accomplish for other reasons.
1) The amount of information a given physical instrument can glean from observing this remotely is limited based, in part, upon the physical dimensions of the instrument. Electronic processing and other techniques frequently misused on TV (like CSI), can NOT add any information to what was actually collected. So a 100x zoom just would not be possible at some wavelengths.
2) When a biological organism evolves an ability, often it pays a penalty - even if that penalty is limited to adapting cells that used to do one job to do another. Human eyes are limited to the frequencies they can see because on Earth, these frequencies provide humans with the best sensory input. Biological organisms pay too high a penalty for the benefits gained when other frequencies are detected. It Costs Too Much
We can side-step the cost penalty to biological organisms by (as you say) developing implant technologies to make up the short comings of our current systems. I suppose this is true.
I also bet we wouldn't see much benefit by detecting other frequencies for the vast majority of people.