I would assume that Tesla VINs (sequence numbers) are assigned on order confirm with no relation to production begin/end dates and that there are occasional skips in the sequence number.
VINs are assigned roughly 6 weeks before production is expected to start. So many 85D orders didn't get VINs until sometime in January. Some still don't have their VINs.
Of course plans change and even when they don't things don't always go according to plan so there's some variation in that. No idea if there are skips, that would take a lot more samples than we get from seeing what people say on the forums to know for sure.
I personally have used VIN assignment to speculate on production. Assuming that you can somewhat guess how many cars they have produced by the VINs they are assigning today. If they're consistent with 6 weeks and the VIN someone is assigned today is going to be made in 6 weeks then you can figure they have 6 weeks worth of production backlogged. There are two problems with this analysis.
1) You need to know their weekly production numbers. We know a little about this since it gets mentioned in investor letters. Their plan was to exit 2014 with 1,000 cars a week. They've also said they want to double production every year. So that would mean they'd need to have a weekly run rate of 2,000 by the end of 2015. However, these production changes are not likely to slowly ramp up but to change in leaps as they make changes to increase their production. For example if Tesla added a 3rd shift (I believe they run 2 now). That would increase their production by a 3rd. That 3rd shift might take some time to ramp up but eventually you'd have a fairly dramatic increase. There's a lot of speculation that they are up to 1,200 vehicles per week over on the investor threads. I've used that number myself largely because they really need to lead ahead of their doubling numbers in order to actually double production in a year. But that number might even be conservative, we just don't know. The most interesting time to do this is when you're close to an end of quarter and sadly that's when you have the least accurate information.
2) It can potentially fail pretty badly if there are major shifts in their plans. If Tesla assigned VINs and then decided not to make a large amount of those cars till more than 6 weeks later then the estimate will be significantly off. These sorts of things may have happened. There were delays with the 85D production start (seems to have gotten going roughly 2 weeks later than expected). They obviously did not idle the plant while that happened and continued to make other RWD vehicles and P85Ds. There is also the fairly common rearranging of production to limit inventory across the end of a quarter. Back in Q3 2014 it was clear that some of this happened at the expense of orders that were already planned. You can try and account for this by executing the estimation strategy and then comparing it to numbers we know. Coming up with an error factor and then applying it to future estimates. But that only works if the errors are fairly normal, which they may not be.
So maybe. If the presumption of 1,200 cars a week is true then they could have done 5,000 cars in January. That puts them on track for 60,000 cars in 2015, which is also a number that a lot of people have thought was possible.
