After listening to the call, and given the rest of the information gathered thus far, this is my take so far.
First, I think Mr. Musk, with all due respect, did not perform particularly well in selling this to TSLA shareholders. That's regrettable. But I believe that the underlying long term trends make sense. I just wish Tesla and Mr. Musk went about this differently.
Next, thanks to
@ntam which provided a link to this article:
SolarCity's solar + storage play on Kauai
The Kauai project is likely the largest Tesla Energy project thus far. It is done in cooperation with SolarCity. It is enough storage to allow Kauai to go completely onto Tesla Energy batteries for their 5 pm to 10pm load. It's 52 MWh of storage with a 13 MW solar array. Here's the kicker...
Kauai will be buying power from this solar + storage system at 14 cents/kWh.
I really, really, need back up sources for this information. According to the article, Kauai will be buying power from a solar + battery array sized for the entire 5 pm to 10 pm main night time load only $0.02 higher than the average cost of electricity in the U.S. I really need some proper analysis of the costs, margins, etc.
Ok... assuming that's true, this is based on technology and cost structures that are about to get much, much cheaper. The Tesla Gigafactory will lower the cost of the storage cells. The Solarcity Gigafactory with Silevo could lower the cost of utility scale PV solar. Even at 14 cent/kWh, the number of places in the world that this solution makes sense is immense. At a 20% drop in that price, it is below the average residential grid cost in the U.S. The demand would be almost incalculable at that point.
So, next year, the Tesla Gigafactory starts to pump out 20700 cells. The initial production runs are probably destined for Tesla Energy products. This requires a new product iteration. Further, if they have the capacity to ship upwards to 10 GWh of TE product next year, at the original $250/kWh TE PowerPack pricing, that's $2.5 billion dollars of revenue. Someone needs to calculate the amount of labor needed to install 10 GWh of TE product. It's likely massive. The skill set to install TE product is basically the same as solar installation, just without the solar part. But of course, it makes sense to also install solar at the same time for many places. Tesla Energy would have to grow really fast to move from producing/selling/installing ~100 kWh of storage product to 5-10 GWh of storage product. And continue on to 50 GWh of storage product in the early 2020's.
So as Tesla designs the next few generations of PowerWall + PowerPack with Tesla Gigafactory production cells, I get that it makes sense to work heavily with Solarcity, including leveraging their grid management software and processes, their sales network, and their installers. And they made the calculation that it makes more sense for Solarcity to be a part of Tesla at that point.
I don't know how much of this is really about residential PPA, which is what really freaks most of us out.
Mr. Musk has been unwilling to provide us with a clear snapshot of what he sees, citing future product plans. And that this is a "no-brainer." However, we do need some specifics to understand what is going on. Getting confirmation of key metrics I have above would be a great start. I still don't see why this shouldn't be better done at the Gigafactory grand opening party next month.