It's a little more complicated than that. Residential Service is a pain for utility companies. Their capital expenditures are amortized over up to 50 years, and rate structures don't reflect their cost structure due to political reasons.
My numbers are ROMs and old, but for a utility to serve a single family home they allocate $2,000 for direct capital expenditures which amortizes out to about $10-15/month. There is another $7-10/month budgeted for maintenance and repairs of the distribution side (sub-transmission:distribution transformer, MV distribution cable plant, poles, and LV transformer), $2/month for billing/collections, $1 for regulated programs, and then $2-3 for base operating profit for these fixed costs. That puts the cost of just having you on the grid at 120/240V in the $25-30 range, without actually doing anything in terms of energy transfer.
[A commercial business in contrast has an amortized capital cost closer to $20 (≤400A, 208V 3ph), but the other costs are the same.]
Your typical residential customer without a battery is going to push ~6x their average demand to the grid for 4 hours a day, and then pull it back at a slower rate for the remainder. That means they can handle about 17% of houses on a transformer having solar before they start having additional costs. For distribution circuits they can handle a slight net export before things get complicated, but you are looking at less than 20% solar penetration. Load variability within those parameters isn't a big deal, as it usually balances out geographically over a distribution circuit (or at least over a sub-transmission circuit).
The best solution is residential battery storage; that can give you up to 50% solar penetration before things go to shit. After that you need some substantial demand-side management or grossly oversized infrastructure (think a per-customer cost closer to $70/month).