First: I'd guess that by light dispersion you mean chromatic aberration. In general refocussing the light will not undo chromatic aberration, but it's possible to specifically design pairs of lenses to (mostly) eliminate it. These are called achromatic doublets.
Second: yes, you need a bandpass filter. It's not possible to transmit everything below 600nm because the glass of the lens will start absorbing in the ultraviolet and shorter. However you can design filters to transmit over a range e.g. 600nm to 400nm. Google for optical bandpass filter to find lots of suppliers selling bandpass filters.
Last: in principle anything will work as a filter, though the range of wavelengths transmitted may not be convenient. For example carbon dioxide filters out infra-red light, which is how it causes global warming. There are a few coloured gases e.g. chlorine, bromine and iodine, but the range of colours available is limited compared to what can be achieved in a solution or solid filter.
Response to comment:
Re the first question, in principle I think you can refocus the light from a prism to a point, or at least to some arbitrarily small area, But I don't know offhand how to do it.
Re the second question (which I can answer :-): the sort of filters used in cameras typically rely either on the optical absorption of some chemical or on a fixed interference system and they can't be changed. When you need a tunable filter you would normally use a diffraction grating. The amount the light is diffracted depends on it's wavelength, so the originally white light is split up into the different wavelenghts. By adjusting the angle of the diffraction grating you can control what wavelength light falls on your experiment. If you've ever used a spectrometer at college this is almost certainly how the spectrometer adjusted the wavelength.
