Hi Niels and jmepler,
first of all, I am totally impaired towards each and every database. I have no shares, no hate nor blind love for one or the other. I am just looking at what is on the 'market' and how it would be improved (in my very personal opinion).
eBird is obviously the biggest and just the amount of users makes it the most reliable for having up-to-date info on certain species at most locations around the world. For me, it's just missing that last bit of the puzzle (GPS) that, in my opinion, is crucial for both users (like me) as for research.
For users, having exact GPS is so much more comfortable when you are in a totally strange place looking for a bird, and for researchers, the exact location enables researchers to link habitat to bird sightings (this is now flawed by the ebird design, only big patterns can be retrieved).
That crucial missing piece of the puzzle is technologically very much available (GPS location is very much the center tool for many smartphone apps), and it should be possible to implement it in the current database infrastructure (I say should, as it could maybe need a whole new design of the database (I am not a database design expert) and/or the visualisation of the ebird map on the website (I fear having flags for every sighting at exact GPS would be too much data to visualize without having problems loading the map).
So I am wondering why it isn't implemented (yet?).
I can agree that eBird is a research tool that doesn't have to be designed for certain users, but at the other hand, like many big applications eBird gets your data for free, and the carrot they are holding is a user-friendly website and system to get an overview of your data, and a look into the whole dataset in order to be inspired to go out birding and looking for birds. It works in two ways. For the (free) data they get daily, and in order to lure more people using their system, they could make some improvements...
@ Niels: when I am at a lake, I input 6 smews, 5 goldeneyes of which 1 male and 4 female, 8 northern pintails, etcetera etcetera. So yes, everything. Some people only input new birds, or one special bird at a location, etcetera. I don't think this is any different between databases: it is still the user (and the ability of the user to detect every single species) who decides if he inputs everything or not.