Flex for the masses? Is Flash Remoting a dead man walking?

I drove out to Emeryville to see Mike Sundermeyer (VP of Product Design at Macromedia) give a talk about Flex at eBig (the East Bay IT Group). The talk was good: while I didn’t learn much new about how to use Flex, but I definitely have more of a sense of where the product is going than I had before.

The second half of the presentation felt like a live version of the Flex mailing list, with Mike being peppered with questions about everything from licensing to future plans for the entire Macromedia product suite. Now obviously, anyone in that position is going to have to bob-and-weave, because certain things they know just aren’t meant to be shared with the general public. But it was possible to at least make intelligent guesses as to what Macromedia is planning on the basis of his answers (or non-answers).
Usual caveats: if someone is a good communicator, they’ll get you to believe whatever it is they want you to believe. So I may have been spun.
1) The standalone flex-to-swf compiler that everyone on the mailing lists has been calling for does seem to be part of the long-term plan for Brady (the Flex IDE). Mike said they’re waiting to make the server product successful first, but they definitely intend to ship a version of Brady that has a compiler built in. Eventually. Now, obviously, this could just be happy talk: he didn’t commit to anything by saying that. But it was a much more positive response to the question than I had been expecting, or that I’ve heard in any previous communications (written or verbal) from Macromedia.
2) I’ve been trying to decide whether to use Flash remoting or xml over http to connect the RIA I’m currently working on to a remote data source. Mike (while definitely avoiding saying anything directly) decided for me with his non-committal answer to a question about the future of Flash remoting. Anyone wanna buy my copy of the Flash Remoting Oreilly book?
3) Macromedia makes a big deal about how quickly they can deploy the latest version of their plugin to a majority of browsers worldwide. Mike brought this up a couple of times, at one point contrasting the 92% adoption rate of Flash 6 with the 40% (!) adoption rate of Windows XP. The reason he brought this up, I think, is that Flex doesn’t work with Flash 6. So in order to code to Flex right now, you have to be extremely confident that Flash 7 (currently at 50% or so browser penetration) will be at 90% by the time you release your product on the unsuspecting masses. Given how long it will probably take to get your team of developers up to speed on Flex, I think that’s probably a semi-safe assumption. ;->