Wouldn't it be nice if he was alluding to the number of Model 3's produced?
Elon Musk on Twitter
"Now at 6400 hats …"
From where I sit, this does not seem to be an Elon who is sleeping on the factory floor in either Reno or Fremont. I feel like he is the type of boss that would be standing directly behind the guy or team that is fixing a specific bottleneck until its fixed and then on the next. Maybe he is just taking time to have a bit of fun on twitter in between, but I get the impression that the ramp is back on track.
I also want to call out a bit the idea that because they produced 1300 less cars in Sept then originally projected and weeks worth of production behind in Aug that they are not 4 weeks behind schedule going into Nov. There are two main reasons that some are looking at this backwards. Even if there are a couple of bottlenecks, improvements continue throughout the entire production ramp including improvements with suppliers. Everything does not stop just because there is a problem with welds on the battery packs. They can build the entire car and have it waiting for a pack for instance. To be clear, I am not saying that is what they are doing, only calling out that even several bottle necks does not stop the seat manufacturer from ramping and other system sin the process form ironing out issues.
The other reason is that this is the bottom of the curve, so you have to look at the entire years target for production and divide the number of cars short by that to find what percent they are actually behind. You can then multiply this by the number of days to find out how many weeks, though you might want stick with days to get how far they are behind the targeted number for production, though this will be exaggerated because the delays did not stop other ramp related process improvements. If I am right, they will actually catch up some later in the ramp because those parts will go a bit more smoothly because they improvement independently of the bottlenecks.
Not having any real data, lets look at a simple scenario. Lets say Tesla is able to production 2000 cars for Oct. If the target was 4000, they would be roughly 6 weeks behind at this point. This would put them ~3063 cars behind (I cheated to make the math simple). If we assume the total production target was 30+100+1500+4000+9000+16000 = 30630. Based on this scenario, 10% of 180 days behind, or 18 days. This equates to a sum total of 6 weeks behind actually dropping to 10% of total production or 18 days of actual production time. Again, this number might be exaggerated based on the theory above of improvements throughout the system even though bottlenecks stopped fully produced cars. We should know over the next couple of months if they can actually exceed our projections for Nov and Dec of roughly 2k/w growing to 4k per week and of course 5K produced (not delivered) in the very last week of the year.
Looking back, it probably would have been much smarter to project the ramp to look like.. 30, 100, 250, 500, 2000x4, 4000x3, 5000 in last week. It would have made it easier to have upside surprise and if Elon really is that confident in the final couple of months, then it probably would have been more accurate.