I figure that the early adoption is a mixture of drayage and defined and shorter routes where the vehicles can be charged at their base of operations where it parks daily (nightly).
That's a slice of the total market, and yet I think a big enough slice to keep manufacturers building trucks for a few years trying to satisfy this demand. That demand is dependent on the fuel cost and consistency being enough lower that not transitioning will be a major factor in the cost of transportation (and I think that it will be - about .20 / mile fuel cost vs. about .50 / mile fuel cost).
The long haul, owner - operators, and other more general use cases will need longer to develop. Primarily due to the need for somebody to build out a significant charging network that can be relied on. I trust Tesla to build that network, but this network is going to be more difficult to build - partly due to the power draw, and also due to the restrictions on where these can be placed (truck stops) due to the rest of the services for drivers and setup they need for handling tractor + trailers (a supercharger in a restaurant or hotel parking lost just won't work for these charger locations).
My simplistic view is it is a choice between 50 Semis a day at Lathrop by December or 500 semis per day at Texas in 2 years time and the answer is .. Why no do both?
I find 300 semi per day by March 2021 anywhere hard to imagine, especially if it involves stamping and paint.
But I also think this 2 stage ramp fits demand well. Tesla may use say up to 1,000 semis internally, or 2-3 months worth of production (allow for slow initial ramp)
IMO large fleets are evaluating small numbers of semis, not because of driver training/adaption but because they want definite evidence of reliability in the field. Things that could go wrong are software bugs/component quality issues. They are taking that conservative approach because it is sensible risk management. They will probably be ready to place bulk orders after around 1-2 years of evaluation..
So again I see a synergy here, by the time Texas is being built Tesla might have firm orders and know the target production volume.
The simple way to think about Megachargers is one Megapack contains enough electricity to fast charge at least 3-4 semis. So I think most Megchargers will contain at least 1 Megapack..
This is probably another reason why the timing of semi production corresponds roughly to battery day, Tesla needs cells for the semis and for Megachargers.
This is also a clue about how Tesla is disrupting the Energy Industry, batteries, VPP, and Autobidder. I have always expected Battery Day will include a fair focus on Energy Storage batteries.
More generally there is a natural synergy between Solar, Energy Storage Batteries, Fast Charging and Autobidder, all are complementary and the combination is disruptive.