Without an ICE motor and drive-chain, why "$15k for Car body, interior, suspension, chassis."
I guess they could get into the $10k range if they cheaped out the interior. I imagine ICE cost is about $1.5-$2k. Transmission is probably another $1k. Various intake/exhaust/engine monitoring components add another $1k. So taking a Base Chevy Cruze (assume $2k profit, which is probably generous) at $17k Subtract $2k engine, $1k transmission, $1k misc, $1k profit, you end up with $11k for the body, chassis, suspension, interior. I bet GM with their volume, has economies of scale where they can build the same car (save for drive train) for $2k less than Tesla can, but I think assembly of an electric car is a little bit cheaper. But that puts Tesla's cost at $13k. Plus I don't think Cruze is really the quality marker that Tesla wants to hit, but honestly a Cruze is just a Cobalt in my mind, even though in real life I know this isn't quite the case.
Body, chassis, suspension, interior have a lot of material and labor costs involved with them. Interiors are way labor intensive, and quality control headache. Suspensions are pretty cheap to produce but require testing after the car is basically fully assembled, and rework is very expensive. Chassis and Body require a good amount of material cost. If Tesla is going Aluminium then you can count on these costs to be double what a steel car costs. And paint isn't cheap. And is a quality control headache.
You could probably drop all my numbers 10-15% and get a workable car. I see going under $40k really a dream for their next model, because Bluestar is going to have to bankroll their scale up. And going 20k-30k units a year to 225,000 (200k + Model S + Model X) units is really not a good idea. The really need to be about 75k-100k units before they try to add on another 200k units. Especially because of all the labor they would need to scale up. Not to mention all the sales and maintenance staff they would have to hire, and the stores they would need to open. I work in manufacturing where it takes 7-12 people to run a line. So adding 2 lines only takes hiring 20-30 people. And generally that can double production. Going from 20k-30k units to 225,000 units a year will require probably 300+ people or more just to manufacture them.
I think Tesla's Bluestar will become a 200,000 unit car that may sell for $30-$35k when it goes to 2.0. I think manufacturing, labor, and sales scaling up will keep Bluestar 1.0 in the $42-$47k range. That and Tesla is going to have to slowly bring down their margins, it will be scary for them to go from 25% to <10% in a single jump.And remember a $45k car is 21% cheaper than a Model S.