I think we can safely assume that if we don't make the 26k deliveries we need to reach guidance, it will not be because of a demand issue. We know that VIN assignments have been happening at a fast enough clip to make the 26k.
From about late July through the end of 3Q16, TSLA was pushing every demand lever they could find to get cars out before Sept 30. We have seen nothing like that this quarter, but they're still maintaining guidance, and the InsideEVs numbers certainly seem to suggest that US deliveries so far this quarter are lackluster to say the least.
There are two main forces I can think of that would yield the data we see.
1) Much more geographic batching than in past quarters. With MX finally unfurling its wings and flying, we know that there are many overseas MX orders to be filled, and so a much greater than normal amount of the quarter's early production could have been for overseas delivery than in past quarters. Additionally, the large demand in China and HK continues to grow, and with a 40% First Registration Tax being imposed on BEVs and PHEVs starting April 1, 2017 in HK, Tesla may have allocated additional vehicles there to capture demand before the tax cuts demand severely. Other overseas regions have similar changes to taxation or incentives in the immediate future as well.
2) AP2 changeover. We know that the factory was expected to be shut down for 2 weeks in 4Q16. What we don't know, is if that 2 weeks has already happened during the changeover or if its happening around christmas time. Makes sense that it would have been done for the AP2 changeover, and maybe we won't see a christmas shutdown. If the AP2 changeover did have some QC issues that slowed the line down longer than the 2 weeks, it could certainly appear in Nov deliveries.
Rumors of buyers having their US delivery dates pushed back would be consistent with either of these, though I admit that the geographic batching was likely well enough defined in advance that it shouldn't have impacted US dates, so I would estimate that it lends some credence to the rumors of AP2 changeover glitches.
The other piece of the puzzle we don't know is how fast the line is capable of operating at max speed. We know we were targeting 2400cars/wk for the end of 4Q16. Can the line go faster than that to compensate for glitches? There must be some upper limit to how fast the line can go, but nobody really knows what it is.