You're only assuming max utility installation, yes would have to do grid upgrade to sustain continuous utilization in those scenarios. I highlight 1500+600+ all flash/blade2 10m throughput to show system is parity with most heavy utility gas model. But for many installations, 100-200kwh grid + 1500kwh battery peak shaving enough to decouple from grid / not require immediate grid upgrade.

1500kwh storage + whatever commercial local grid plugin (i.e. 100-200kW drip) with battery peak will suffice in many current market scenarios. Have to factor in actually blade2 adoption timeline and charging habit etc, i.e. 1500kwh+100kwh/200kwh = 4000-6000kwh which is enough for 40-100 cars (not all going from 0-100, avg ~50-60kwh topup charge). For most markets this is gas pump speed for blade2s + flash, or ubiquitous charging without major grid overhaul for legacy. Will take years for blade2 saturation so that is years of infra/leadtime buffer, or alternatively just slap more piles + battery to simply reduce grid contribution, but under CURRENT market conditions, this is drop in installation for immediate value, with multi year lead time to upgrade grid if required. THE MARKET REALITY for years is 2x1MW/stall all the time is going to be edge case, and something software (i.e. PRC already have energy rate arbitrage queueing) will have to address as they become more common / battery buffer cannot keep up with. Until then the important infra consideration is installation is 100-200kwh spread over multiple likely preexisting commercial level transformers (which is months leadtime hack vs years utility for less infra competent regions). Napkin math is ~10 (i.e. 20 hoses) of these will match typical western neighbourhood gas station throughput. This not considering markets where home/residential BESS will shift even more charging to night curve, i.e. we should expect less 0-97 behavior and more topup demand.

Yes execution difference is PRC transformer / utility lead times is in months to hook up the full 600kWh+, but for everyone else the system allows them to hedge with battery in lieu of immediate utility scale grid upgrades. The battery buffer realistically defers major upgrades to enable installation now, because even if upgrade required, it would be commercial vs utility scale in terms of paperwork and leadtime. When I bring up KFC, I'm not saying they're using KFC power, I'm emphasising system works on commercial level power hookup not utility level, i.e. commercial power during night/lulls + battery peak shaving + small spatial requirements + drop in WITHOUT upgrade opens up much more micro charging sites, instead of a charge hub with utility hookup, can spread piles over dozen restaurants with local grid without major upgrades for years until blade2 or more hungry packs reaches saturation. KFC doesn't need 2MW to spare, the immediate commercial grid just needs 100-200kwh to spare, i.e. enough spare capacity for another walk in freezer, most commercial sites have that overhead and battery peak shaves / recharges / balance typically synergize with with business activity.