I read some of the dicussion here (didn't read every single post though
). I think that adaptive music is totally possible today. With our company we actually do it today. Although it is challenging from a musical point of view as well for the audio engine used, we can produce adaptive soundtracks. Modern audio engines like Audiokinetic's Wwise support us in cross-fading musical ideas on the beat without being noticed by the player. All these transitions or mixing of different sound layers can happen in reaction to any in-game event. Sure it is very challenging to compose music for interactivity and for such an engine - but it is also very rewarding in terms of emotional depth for the player. It is actually this interactivity that differentiates movie composers from game composers. I'm a bit sad that most of the game music today is produced by movie soundtrack makers. Sure, they know how to support emotions musically, but they usually have totally no grasp of interactivity. And don't even try to ask them to script some in-game events to trigger audio...
Composing for adaptive audio is a new challenge and we hope that music and game schools will recognize this and integrate adaptive composing into their curriculum.