The way the blogpost details the process of using the spreadsheet is nice - giving one simple example and then scaling the number of stocks clarifies when this sort of thing would be useful.
However, it would have been nice to have a bit of explanation/background on how the PCA algorithm actually works.
TechCrunch basically paraphrased this article (http://techcrunch.com/2012/07/31/the-mechanics-of-a-small-ac...). It's nice that Stypi is getting more press, but is it weird that TechCrunch kept the same format and only mildly changed the wording of the original post?
This post is right on the money. The very small chance that someone will "steal your idea" based on a short pitch is completely outweighed by the much larger chance that the listener will offer useful feedback or that your conversation will eventually have other beneficial results.