Red Hat and Fedora
Jon Udell has posted a partial transcription of Michael Tiemann's talk at OSCON. From this corporate perspective, it looks like the strategy is dead-on. They had to slow the release cycle so that they could provide a common target for the key application providers. In a big corporation, there's always pressure to look for complete platforms rather than a niche ones. The attitude is, "Who wants to support yet another OS?" In fact, our first non-appliance production implementation of Linux will be an Oracle database. Without this sort of application support, we still wouldn't be doing anything "official" with Linux.
As for Fedora, it sounds like a logical response to the (probably necessary) overcorrection that Red Hat made to a corporate focus. It sounds like they are aligning their resources with their markets -- 80% on enterprise and 20% on early adopters.