There’s a longer post coming on this (been busy recently) but the Oct-Jan numbers are not seasonality. Demand hit a peak exceeding 6K units in Oct and practically hasn’t come down.
What’s happened is that we had the largest decommit ever by 1 electronics partner in end-Oct. Cell supply has been also pretty erratic so had to accept massive line-downs Nov onwards to balance the constricted supply. We even slowed down on new EC openings because customer lead times had sky-rocketed and it was getting difficult to manage old order wait times.
In the meantime, decided to bite the bullet and pay crazy spot prices for some of these components and resume production at a low level this month. On the bright side, new orders went in the system for all suppliers starting July and with lead times of ~40 weeks for bottleneck components we should start seeing results starting April. Also much better relationships with partners across the board now.
Massive ramp-up ahead, demand stopped being a problem a long time back - it’s mostly about managing supply chain well now if you have a good product. Hence, the plant expansion. The minute we cross 13K units/month the current plant will be 100% utilized and that number (13K) looks barely 30 ECs away. All of them are in the pipeline in the next 5 months.
Charging infra - yeah, donno what I was smoking (wasn’t sitting very close to ops back then) when I said 6500 units years back. Easy to announce #s when you haven’t built much yet or think that it’s totally sensible to burn 100s of millions $ pre-revenue 
It’s gone through it’s own design hell (metrics look stable only now) and will expand. But will take ~2+ years to get to that large a number still.