I think there is room for such a project to co-exist with wikipedia, but I think you're vastly underestimating the human obstacles to creating something that is not just a weak half-clone of WP content.
I don't think I am underestimating these challenges - in fact they are one of my main worries, but will be sorted out when they become more urgent.
The chief problem I see is to get some sort of cross-discipline cooperation. For instance, FishBase has made a good start with fish, but the database entries for beetles or plants will be mostly different; sure, there is some sharing, but the coleopterists' approach to taxonomy is (seemingly :-) ) rather more chaotic, and, well, the botanists have seven different ways just to define the concept of "species"...
And so do Zoologist and there is a great argument going on about how many KINGDOMS apart from animals and plants there might be - but again: There are plenty of contents we will be able to put together even if the taxonomic background of a particular species is unclear. If fact, the wiki approach is perfectly suitable for a dynamic science such as taxonomy.
So you're talking about pushing all of these specialists into a single framework, and if they don't fit well, they're not going to participate.
They will: An expert on bats of Panama will not care about whether archaebacteria have thir own kingdom or not - as long as the bats' features are worked out properly. Every branch of the taxonomic tree has its own leaves, in paper-publications as well as it will have to be in wikispecies.
So before talking to developers about software, you need to talk with people in different areas and get an agreement in principle. Could the FishBase guys sign on to import their data into a cross-phylum project?
I don't agree with that. Fishbase is fairly scientific and partly commercial. It is not open access as wiki projects are. To me it is the best species directory on the web; still, I don't know if fishbase will co-operate, I will give it a shot as soon as I got technical support. They do co-operate with Species2000, so there's hope...
And finally, what content would this have that is not just as appropriate for WP, and do the WPers agree with setting that boundary? For instance, the full list of papers reporting every sighting of a species of plant seems too detailed for WP, but I could imagine a parallel set of "dig deeper" articles that go all out on that sort of thing.
Your example is a good one. Also, a determination key might be useful and generally a lot of details that are not relevant in an encyclopedia. Schematic drawings. Keys for text-only browsers that can be read on PDAs for field work. A range of things that WP would not want to provide.
Benedikt
PS: Anybody keen on helping out with the tech-part? Setting up a basic structure?