One comment I've seen repeatedly is that EVSE pay ToU rates plus "burst demand" rates. Supposedly even if commercial electric rates are 2-3c/kWh, when burst demand pricing is applied, the cost per kWH can exceed $1.00
This seems like a somewhat necessary evil, given the unique demands of EVSE.
If the burst demand charges put rates as expensive as $1.00/kWH, it could/should encourage onsite battery storage combined with solar canopies (or even wind). The batteries could be topped off at night with the utility providing more attractive rates, while the solar panels do what they can during the daytime to keep the batteries charged. Then through battery management software, the onsite battery storage could be "trickle charged" throughout the day to reduce these burst demand charges.
Demand charges are not charged by the kWh, but by the kW, and they are usually far higher than $1/kW (in my area, over $10/kW for commercial service).
So what does it mean to charge a per kW demand charge? It basically means that over a given month, the highest peak demand you use from the grid ends up being your demand charge for the month. For example, if you have a Supercharger site with 750kW service, and you have a full site pulling that maximum 750kW all at once, you will have $7500 in demand charges for the month at $10/kW demand charge, in addition to the per kWh electricity charge. Of course you have to spread that demand charge out among all your monthly users (along with equipment charges, real estate charges, insurance, back-end billing and other overhead), so you work out how many kWh you dispense in a typical month and tack on a per kWh surcharge to what you charge customers per kWh to cover the demand charge.
For a busy 8-stall Supercharger site, it's probably not that bad. Maybe you have 3500 sessions per month with an average of 50kWh per session delivered (175,000kWh). The $7500 demand charge "only" adds about $0.043 per kWh. Of course if you were only paying $0.06/kWh, that's a pretty big increase percentage wise. And you've got all that other overhead too.
But if the Supercharger is generally not "busy" and you only have 1500 sessions over the course of the month, except for that one hour of that one day when it's fully loaded, then bam, your demand charge overhead jumps to $0.10 per kWh!
It gets pretty bad for a one- or two-unit site though, as we see with CCS/CHAdeMO sites. Those tend to be lower power of course, but some are 125 kW units. The demand charge for that might be $1500 (assume it pulls 150kW peak from the grid--after all it has its own cooling equipment that takes power as well), but with a single stall you might only have 15 sessions per day (many non-Tesla vehicles charge at 50-80kW and charge sessions are longer on average), or 450 sessions per month, and let's call it an average of 40kWh per session (smaller average battery size) for only 18,000kWh dispensed. Now your demand charge overhead is running $0.083 per kWh!
These are only back-of-the-napkin figures, but it does illustrate how demand charges work and how they need to be accounted for, and why as much as we may not like it as consumers, business model wise, an huge Supercharger site with many empty stalls is generally not a good thing, cost wise, for Tesla.