The intermittency is sufficiently predictable that we know the overbuild requirement well enough for a variety of possible solutions, including cheapest overall, or least storage requirements. I tend to go for x10 in armchair discussions like these because that's the level needed given the current global average power factor of PV.
> Are they getting cheaper enough? Including all the required overbuilding and all the (yet non-existent) grid-scale storage that is required for renewable energy?
Yes.
> grid storage required at these scales is literally nowhere to be found
The fact I can say the same for the uranium mines needed for sufficient nuclear power, isn't an argument against nuclear power.
Why?
Because we don't have a current need for those mines.
Why is this a useful comparison?
Because until we decide we want to electrify transport (it's mainly about that not about PV vs. Wind vs. Hydro), there hasn't been demand for that many battery factories.
But in both cases, we know what to do, and how to do it.