As a long-time ColdFusion developer, this announcement is one more point of frustration with the fact that ColdFusion seems to miss out on the larger audience exposure that its siblings seem to be enjoying. While this announcement seems to be in the context of promoting "Web 2.0" technologies (a Tim O'Reilly-coined term that I absolutely loathe, by the way), this seems like the type of agreement where letting ColdFusion ride the coattails of the Adobe Flex 2 marketing strategy could be a win-win for both sides.
Why? Well, you can't do a whole lot with Adobe Flex 2 if you don't have a server-side technology to do things like database querying, authorization/authentication, etc. It's a presentation layer technology (the "View" in the Model-View-Controller pattern for those wishing to be buzzword-compliant), and Adobe has been preaching for months now that the single easiest way to integrate Adobe Flex 2 with your back-end (the "Model") is via ColdFusion (in this case, the "Controller"). So why not add ColdFusion MX into this agreement and subsequently provide resources for showing just how smoothely integrated the two technologies are. You'd have the expertise of the Adobe ColdFusion team providing high-quality examples while the O'Reilly team provides the author/publishing/conference resources to expose this seamless integration to the masses. In return, O'Reilly gets an extremely passionate ColdFusion development community to support their efforts.
You almost literally can't find a competent developer who hasn't heard of or used an O'Reilly-published book -- they're everywhere and are usually excellent resources. While I know that they decided a while back not to publish a ColdFusion MX 7 book due to sluggish sales of the 6.1 version, there could have been alternative resources published with this agreement that might have focused even a small part of the spotlight on the power and ease of developing with ColdFusion (i.e., O'Reilly-backed articles, samples, etc.).
Adobe currently seems to have their marketing engine squarely focused on Adobe Flex 2 and ActionScript 3, as well they should. I just hope that the same zeal with which they're pursuing the growing rich internet application development community is equalled or surpassed with next year's ColdFusion 8, aka Scorpio, release.
While Adobe/MM has been talking up the sales of CF7 quite a bit over the past year, the gains appear to be incremental rather than exponential. I don't think that even the new features rumored for CF8 are going to cause the droves of .NET or Java developers (or the 10 guys that get paid to develop with Ruby on Rails) to suddenly jump ship. Adobe's only chance to really pump up CF sales is to hitch its wagon to a star: Flex.
Pushing CF as the premier backend for Flex could do a lot to accomplish that but you still need to sell people on (1) the need for a rich UI, and (2) that the Flash platform is the best way to do it. But once you win the front-end battle, there's the still the chance that people will drop coin on FDS which marginalizes to some extent the value that CF can provide on the backend.
I hate the term "Web 2.0" but the more I see it mentioned on Digg, the more I realize that to developers it means 2 things: Rich UI (usually via AJAX) and simplified, rapid development (Ruby on Rails). Maybe another way to boost CF sales would be to tightly integrate Flex and CF into a "single stack" environment much like RoR. Maybe even bundle FB with CF with an attractive pricepoint...