Ah, yes, that makes sense. At the back of my mind I was thinking of the recent "OMG MANTIS SHRIMP CAN SEE 12 GAZILLION COLORS" meme and I realized that no, they just have receptors for more different wavelengths than we do, that says nothing about their color perception, or whether they use signals generated from the respective rods and cones in the same way we do, etc etc etc. Looks like you got a grasp on the fact that it's low level stuff that we're comparing here, so please do carry on 
Fair enough!
In that case, all neuronal populations must retain their respective identities in the face of near-constant environmental change, (a feature that affects all sensory & 'routing' systems). Such homeostatic strategies can be aimed at different levels, from specific subsets of cells or to local and global networks; these strategies have the same aims in all organisms with complex nervous systems, though do comprise different components and structure.
Further similarities may be found in the basic architecture, where the structure/function relationship of certain networks can be linked directly to a physical phenomenon. Take motion detection for an example, where in the fruit fly a Reichardt-type correlator serves to compute the difference between visual occurrences at two points on the retina, thus providing the basis for local interpretation of movement. As mentioned previously, this faculty is reflected in the architecture of the early fly visual system, with 7 adjacent & pooled ommatidial inputs providing the local detection array for such a computation, and downstream layers of the visual brain providing the relevant multiplication and subtraction events. This system forms the basis for a simple motion detector and this is also the basis for more complex models such as Barlow/Levick, some form of which has been proposed to underpin motion detection in mammalian systems.
Neuronal solutions to ecologically-relevant problems, across organisms that share similar environments, will often share a basic structure; though this will obviously diversify and become specialised in accordance with an organisms specific requirements, the modes of regulation and basic architecture can be used as a reference point.
Furthermore, knowledge of the ways in which neuronal systems self-regulate and communicate is not only useful for human/non-human comparison, but is the basis for machine learning and the development of computational devices for many purposes. And all of these phenomena, in combination, help us to build a more detailed picture of how nervous systems work.
p.s. I kind of feel that this might sound like some waffly bull___, so forgive me if it does!
