That's probably due to the absence of an effective AQM as the bottleneck, rather than the transport protocols being unfair. See https://datatracker.ietf.org/doc

That's probably due to the absence of an effective AQM as the bottleneck, rather than the transport protocols being unfair. See https://datatracker.ietf.org/doc/html/rfc7567 https://datatracker.ietf.org/doc/html/rfc7567 for the current best practice