I am sorry for my relative silence: I've been whooping it up in Legoland
Colour is a big topic, but it is important to realise that there is a fundamental difference between interference colours such as those found in Lippmann emulsions and thin films, and the colours of colloidal silver which are related to excitations of the electrons in the silver particles, and are thus more like the colours of pigments or dyes.
Interference colours can be very intense and pure, but they generally only work over a limited range of angles. Exceptions exist which work over wider angles - the canonical example is the pure blue of the Morpho butterfly wing - but as I understand it Lippmann emulsions only display their correct colours when viewed head-on.
I think holographic emulsions are today's direct descendents of Lippmann's work: they perform a similar task of preserving interference fringes in the thickness of the emulsion by laying down correctly spaced lamellea of silver.
In metallic materials where the electrons are free to move you can create compression waves in the sea of free electrons which are analgous to sound waves in a gas. These are the so-called 'plasmons' which crop up as buzzwords in any modern discussion of the optical properties of small metallic structures. Because the wave changes the local charge density, it has an associated electromagnetic field, and in the right circumstances you can couple that field to the freely propagating plane waves of which light is made.
It is these plasmon modes which give colloidal silver its colour. Other metals work too: gold for example turns red and then a gorgeous deep blue as you make smaller and smaller particles. Most experiments are done on noble metals because the large surface area to volume ratio of the particles makes them very susceptible to corrosion and oxidation, but in principle aluminium and iron will display similar effects.
The plasmon frequencies are acutely sensitive to the particle's size, shape and local environment. As I said in my earlier post, nearby particles can radically change the way that a particular particle scatters light, but the properties of the surrounding medium can have a similarly dramatic effect: changes in its refractive index or its pH will often lead to changes in colour.
This sensitivity is one reason for the widespread interest in applications of plasmonic structures. For example, you can already buy blood sensors where the binding of a protein to a gold film changes the plasmon frequency and alerts a monitoring circuit. It is also the reason that plasmonics seems to promise a tempting wide-tunability across the optical spectrum.
The problem for photographic applications is that the same sensitivity makes the colour production unreliable, especially with conventional home darkroom levels of control and repeatability. If time, temperature, humidity, and chemical concentations have to be controlled to 0.001% then this isn't a practical photographic technique, even if the same science ends up being used industrially to make sensors or the next generation of LEDs.
There are of course other mechanisms to make nanostructures more coloured than their bulk analogues. Mie scattering can be highly wavelength and orientation dependent, and the absorbtion that turns large chunks of material grey does not have a chance to work in small systems. Nanowires made from semiconductors show the most beautiful colourations, even in the absence of interference or plasmon effects.
I should stress that I am not a real plasmonics expert, although I do work with a few, and I have very little relevant experience of the photographic aspects of these phenomena. I very much welcome further discussion, and the valuable practical input from those who have actually got their hands dirty in a real way.
Final point: the early colour I was thinking of was bare AgCl on a paper support, as used by spectroscopists to investigate projected solar and other spectra by measuring the pyrolitic or photolitic darkening of the halide when exposed to a spectrum from a prism. Herschel reports on some of these colours in his huge catch-all paper in the Royal Society's Phil. Trans. from 1840 (the same paper where he reports the use of "hyposulphites" as a fixer), but I am sure many other early workers noticed and commented upon them. These are unlikely to be caused by an interference effect, and are most likely plasmonic colours.
Struan