His point is that brotli's advantages aren't big enough to justify it until it becomes ubiquitous. Compression improvements of less than 10% just don't justify

His point is that brotli's advantages aren't big enough to justify it until it becomes ubiquitous. Compression improvements of less than 10% just don't justify much effort to get ahead of the curve (in most situations).