I guess that explains my recent question of why the MPEG side of things stopped publishing comparisons to VP9 after it actually got released and instead continues to point to studies they did from random git commits, 6 months before the first (and even then explicitly called out as beta) release of the encoder.

Be interesting to see the results from this comparison, as they seem to generally be a respected third party in this field:

http://compression.ru/video/codec_comparison/call_for_codecs...