Really enjoying the article so far! I wanted to point out a minor typo I just noticed in the “CDN adoption and usage” section:
This slight variance is likely do to the small continued use of mobile specific web pages (“mDot”), which more frequently use a CDN.
(Edit) and another small one:
Some CDNs support
post-checkas a method to allow a resource to [be] refreshed when it goes stale, andpre-checkas amaxageequivalent.