Hmmm, good attempt at an analysis, but your assumption is wrong that the two things are related (Tesla negotiated rates and planned energy generation to cover it 100%). They are not. Well, mostly not over time". If you lived here in Sparks or in Nevada and understood the way business politics and government incentives work, and the long range Tesla plan, then it would be more clear as to why Tesla negotiates such things. Your assumption is true for Apple with the data center and similar companies moving here, but only minimally true for Tesla. Your energy consumption estimates barely cover the first phase of implementation. Full implementation of the Gigaplants by 2022 is underestimated by a factor of more than 5 !
This is analysis of the Nevada Gigafactory. As far as I know it is not planned to produce more than 105 GWh of cells and 150 GWh of battery packs per year. We need to do an analysis of each planned Gigafactory to check their feasibility in terms of onsite renewable energy, but at this point we don't know the details about the other Gigafactories. If Tesla is serious about being powered by RE, it needs to be buying land in the region to build solar farms, wind farms or geothermal plants, since it currently doesn't have enough land or it needs to be negotiating with third parties to build renewable energy plants.
At this point, the burden of proof lies with Telsa, because it is making claims that don't sound reasonable. They are currently planning a 70 MW solar array on the roof of the Gigafactory, which at a 20% capacity factor means just 14 MW or 122.6 GWh per year.
I estimated 100 -150 MJ per kWh of battery cells, but maybe the Tesla/Panasonic engineers have done something miraculous and have figured out how to use only 30 MJ of energy per kWh of battery cells and 5 MJ of energy per kWh of battery pack in the Gigafactory. Even then, the Gigafactory will need 1083 GWh per year or 124 MW to operate. It will be extremely difficult to generate that much energy onsite, even with geothermal and it takes a minimum of 3 years to build a geothermal plant.
I can't even start on the issues surrounding "fighting for public policies to reduce the number of vehicles on the road." If you haven't looked yet, you may want to research the massive issues in Beijing China with their policies to do this by having a lottery to get a license plate! Lots of negative effects, but few positive ones. Their real issue is the number of coal fired power plants they build.
Many societies have fewer private vehicles per capita than the US without significant loss in quality of life. Many cities have reduced the traffic in their downtown streets. Look at the public policies in Copenhagen, Oslo, London, Portland, Vancouver, etc. I live in La Paz, Bolivia where private vehicles are prohibited from traveling on certain downtown streets during week days during high traffic hours. The policy has worked for the last decade.
China is a very bad example of public policy because the national Chinese government has been encouraging more production of private autos at the same time that city governments were trying to cut the use of private autos. China has increased vehicle production 17% per year since 1999 and many Chinese cities including Beijing prohibited the use of bicycles on busy streets, so it is hardly a good test case for what is possible.
Redesigning American cities for public transport with new zoning laws, greater urban density, urban infill, etc. will be a long-term effort, but so will switching to 100% renewable energy and switching to 100% electrified transport.
I haven't seen any studies about the energy and GHG emissions of making sensors. I would guess that they are high, but I really don't know.
