Maybe I'm drawing too much from my own experience whereas you have more data, but I certainly saw traffic drop a lot around Christmas and not return in January, coupled with much lower eCPM figures. I was generally making about 1/3rd of the return pre-christmas.
On your graph of application success by week of launch, the 4 week moving average seems to decline most significantly from early January to early February, so doesn't that indicate something similar? Or am I reading it wrong?
Great research and write-up - I'd been sensing the same too, but the figures tell it all. My own opinion: the decline started early January when installs and page views dropped off as the pre-Christmas goldrush came to an end. As a result application revenues tailed off and this is filtering through the developer community. I don't know how much Facebook's attitude to applications has hurt, but that and an uncertain future can't have helped.
Maps mashup maps mashup takes it one stage further by using derivative content made from more derivative content. Okay, it's a bit tongue in cheek, but lets face it the idea of showing things on a map isn't exactly revolutionary. Google have just made it easy (for now) by offering a free and usable interface, but it's not a new business model.