The subject says it all: is AIR something that will be around for the next 2-4 years and not just a cool developer toy? Actually, I think so - but currently I need some real good selling points for a project that could be done with AIR but the customer is rather sceptical about the technology and the general roadmap.
The project is a strategic enterprise project so the customer really needs some kind of safety for the planned investment and especially some indications that AIR won't dropped or discontinued in the next 2 years (which is a valid point I think).
What pro-AIR arguments do you know?
Dirk, I've developed four AIR applications for the enterprise (including an application for Avaya and Seagate) and there are a bunch of great reasons to take the jump. Granted, Adobe hasn't done a great job of making those selling points well known. I think part of the problem is that the reasons are very specific to the project.
The tools we've been working on have been largely for the salesforce. Since I don't know your particular situation I'll give you the general breakdown as we see it.
1. Online/Offline Usage. The salesperson may or may not have an internet connection when they are on-site. So the salesperson doesn't have to be worried about losing an internet connection in the middle of a big sale.
2. IP protection. Large enterprises invest considerable amounts of capital into these applications. They don't want to see that investment walk out the door if a salesperson quits. You can closely control access to the tool with some form of licensing server integration. If they're coming from a less refined tool like Excel, then this point should ring very true. However, if they're switching from a traditional web application then its a moot point since they should already have a user system built in.
3. A Branded Experience. AIR applications allow you to bring the branded experience of a web app to the desktop. This is particularly relevant for larger organizations (not as much for small and mid size).
4. Reduced Server Costs. AIR (and Google Gears for that matter) are great on bandwidth and server performance. Remove all the extraneous "view" related requests and just transfer the required data.
More so than selling the client make sure using AIR is a suitable fit for you. Pushing out updates to installed applications and migrating services appropriately can be a lot more tricky that your initial assessment. Web applications are great in that you never have to be worried about the client's "version" being out of sync with the server's "version".
Whatever you decide to do, I wish you the best.