@io @nota @vidister could have calculated 1,77/2 for the same result.

It's not even grouped by service or something. Who knows if >88 % of that traffic is just a single noisy IPv6 service