Let me give a bit of perspective. I joined Afrihost at the beginning of the month (or rather the middle once we sorted out some Openserve issues). The reason I moved ISPs was because my previous provider had been hammered for close on a month by DDoS attacks - similar pattern, waves that tended to converge during heavy usage periods, probably because the existing traffic made it an easier target.
My previous provider was smaller than Afrihost, and they couldn't manage to keep international up at all during the attacks. They did everything they could, but it's a resource-heavy problem. In the end I left because there was no end in sight, but I don't blame them for it.
Liquid telecom is a larger player, with attendant resources available. You can see that in the fact that they're managing to keep things mostly rolling. Yes, it's tremendously frustrating for those of us who use latency-dependent applications like gaming, but we're at least not seeing total failure since the first wave.
tl;dr: it could be much worse, and I think you should re-evaluate the degree of responsibility that you're assigning on the provider side here.
Thanks for this. Totally agree - the guys at Liquid are doing the very best that can be done.

