This is exactly the algorithm we developed at LogNormal (now part of Akamai) 10 years ago for doing fast, low-memory percentiles on large datasets.

It's implemented in this Node library: https://github.com/bluesmoon/node-faststats

Side note: I wish everyone would stop using the term Average to refer to the Arithmetic mean. "Average" just means some statistic used to summarize a dataset. It could be the Arithmetic Mean, Median, Mode(s), Geometric Mean, Harmonic Mean, or any of a bunch of other statistics. We're stuck with AVG because that's the function used by early databases and Lotus 123.