Can someone fill me in on the hate for Reuters/Ipsos polls? They don't produce "consistent" results, but that's not a mark of a bad pollster at all.
They are an online pollster using a panel (partially self-selected). Online polls are generally not the best.
But, on top of that...
You are right that having some variation in results is not normally the mark of a bad pollster, but with an online panel you should have much less variation than with a phone poll, and you are not subject to response bias in the same way.
So the fact that they have had so much volatility despite using an online panel is an indication of junkiness in that case, whereas it wouldn't be so much for a phone pollster.