I'm not sure how we'll know soon enough, considering your parenthetical "other than the numbers". Let's say a bunch of Ps get delivered before my R140. We don't learn anything. Unless 2,000 Ps get delivered before my R140, I got a big benefit from my R number. When I made my reservation, I had a choice of just taking a normal production number, which was around P2000, or getting the sponsorship of my Roadster-owning friend, and getting R140. Your "other than their numbers" doesn't really fit in.
I’ve been thinking about the prioritization of Rs and Ps, and the logic behind what TSLA is doing. Let me put on a manufacturing hat on for a moment and hypothesize ...
TSLA is mixing the Rs and Ps, within reason, to allow them to create batches of similar configurations. Mixing the Rs and Ps allows the factory to increase its overall output, at a given consistency (quality), over doing strictly Rs or Ps. Also, mixing the Rs and Ps allows more time for each group to finalize their order. TSLA can ‘pull’ Rs forward into the production mix, while keeping the Ps in close sequence. This may give some benefit to Roadster Friends and Family.
The ultimate goal of a manufacturing system is consistency (quality = low variability), low cost, flexibility and speed (fast).
When starting a manufacturing operation, especially for TSLA, the focus is on quality. Then speed, then flexibility. Low cost is a function of volume and the other components all working together. Of course, gross margin is, to some extent, a function of the sales process and the ability to sell inherently higher margin vehicles.
Initially, to focus on quality and speed, it makes sense to have ‘batches’ of similar configurations. The flexibility comes later, when every vehicle is configured differently and the manufacturing line operates at speed producing, for example, different color vehicles with different options and battery packs, one after the other.
Having a larger pool of similar configured vehicles allows for batching of vehicles so fine tuning the manufacturing process is easier while production can be faster (than otherwise). GeorgeB referred to this in his email when he stated, ‘In some weeks it meant we had to reach forward in the sequence order to find cars that were not impacted by a particular décor or option, and in some cases the absence of a particular décor or option pushed cars back.’
Also, having insight into the potential ‘pool’ with as much confidence as possible simplifies the batching. The information system supporting the manufacturing is evolving at the same time.
With that as a background, having pools of Rs and Ps, with as many confirmed (locked) configurations enables TSLA to create similar batches taking vehicles from both pools. If there is a configuration of Rs that coincide with Ps then the two can be combined in one batch. Having a two week delivery window determines the size of the batches, within reason.
The prioritization of the batches is influenced by multiple factors and, all things being equal, the highest margin vehicles may be produced first.