There is an obvious traffic jam happening on Intel's mobile roadmap. Everything has seen similar delays, and until they can sort it out, I have little hope of seeing Broxton on time. I can assume that Intel will eventually figure out how to build a leadership SOC under 5W of power (best-in-class CPU, GPU, ISP, LTE, etc.) - and eventually get it close enough to the process launch to have an advantage over their competition. But there's just too much evidence right now to suggest a lineage of poorly planned products, which tends to take years to unwind.
I think the handwriting is on the wall. Better for you to acknowledge it now, than to be disappointed later. If you don't believe Intel has a viable business outside of Broxton's success, then you should sell now. On the other hand, if you think the PC business will stabilize, then Intel will continue to make multi-billion dollars in margin, and Broxton's 2015 timing won't be a hinge factor at all.