Measure of how "concentrated" or "scattered" values are

[Dr.Peterson]

Interesting, instead of expanding the lists you do a cumulative sum, might work better in terms of performance.
My data ranges from 1-53 (weeks) for X. I suppose you would recommend Std Dev then, rather than IQR?

Either might work for that kind of data, particularly since I don't think you need precision. You can try both and decide.