Well it doesn't prove anything other than that both numbers are within the range of what 85D and S85 can get, which makes the EPA numbers entirely plausible (rather than impossible). And while the 85% is at 120V, it was also in a 2013 S85 with updated software with sleep.
Look let me be clear. I don't doubt there has been some improvement. I doubt the improvement is from 85% to 92% efficiency (in well controlled tests). The efficiency differences we're talking about here as seen in anecdotal tests are probably well within the range of normal charging efficiency differences based on climate, equipment differences, current differences, etc...
The thread you're pulling from also clearing shows that he was removing vampire drain. So that's the pure charging efficiency. The EPA test procedure is obviously going to include some vampire drain since they let the car sit. Your theory suggests that a 2012 car was doing better than that 2013 car with the improved software. Because it was hitting 85% after the vampire by your calculation. Of course the 2013 car was hitting 87% charging efficiency on 240V/40A. So maybe the difference there is just the charging rate differences in efficiency. I'll get to that later.
I don't see it as sinister as that. I think in general if the changes result in higher efficiency and you choose not to retest, that will not sound the alarm for EPA (as the EPA likes the results being as conservative as possible). So doing so in that situation matches both spirit and letter of the regulations. Your Ford example was the opposite: they knew the numbers would be lower if they retested so they chose not to (officially I know they said their engineers weren't aware of the aero efficiency differences between the two body styles, but that was taken to be BS by most people). After all, people don't sue because of getting better than the EPA numbers, they sue when they get worse than EPA numbers.
That's only viewing things form the perspective of an S85 buyer. Consider it from the perspective of an 85D buyer who spent more money on the idea that they're buying a more efficient car. If the amount of efficiency improvement between an S85 and 85D is less than provided in the EPA numbers that's going to make people upset.
Not bothering to update the numbers when all the cars had the same EPA economy information and all of them were presumably improved by the same amount doesn't do much. But once you're providing some of your models with updated ratings and some without and you're actually advertising the car to some degree on the improved efficiency. I do think that rises to an actionable situation.
So yes the Ford situation was different. Clearly people in that case were buying cars and getting less than the window sticker. In this case, presuming your theory, nobody is getting a car with less efficiency. But it is distorting the purchasing decisions in the market place and doing so in a way the encourages people to give Tesla more money. I'd also say this is somewhat tilted against Tesla since for all practical purposes their cars are just competing against each other.
Tesla may have not bothered to update the S85 rating because they don't expect to be selling it for much longer. Which would help mitigate this issue.
SAE standard specifies testing with the highest AC power available. The EPA site lists the following under specs:
2012 S85:
12 hrs at 240V
2015 S85:
12 hrs at 240V
(standard charger)
4.75 hrs at 240V
(80 amp dual charger)
I'm pretty sure an optional EVSE doesn't matter, since no EV other than the Tesla comes with a standard 240V EVSE, but the standard still says to test at highest power available (don't have time now, but will dig up exact wording when I have a chance).
So it's entirely possible/valid for S85 to be tested at 240V (or even 120V) and for 2015 S85 to be tested with HPWC.
Just to be clear, the SAE standard doesn't specify how to come up with those charging rate numbers that go on the window sticker. As best as I can tell we have no idea how they come up with those. The old stickers were practically useless since they didn't even mention the dual charger bit. It's obvious that 12 hours at 240V is not 240V at 40A, so that's not maximum charging.
The SAE standard is terribly vague on the charging stuff. But it does specify that you can only use charging equipment that's included with the car. In my opinion that excludes the HPWC and requires the use of the UMC. I'd like to think that the interpretation of the standard means that they should be testing at 240V/40A. But even though the charging rate numbers aren't necessarily based on the SAE standard tests, the 12 hour rate gives me pause on accepting that.
The standard is far from clear cut. Even then it's pretty clear that the EPA has authority to permit deviations form the test procedure based on their own decisions. So without specific information about how these tests are conducted (not in general, but for the specific results) it's hard to know what the heck they did.
And putting aside this whole charging efficiency argument, how do you account for the fact the range numbers of 265 vs 270 showing only a 2% efficiency difference, while MPGe combined shows a 11% difference? Are you suggesting the usable capacity went down for the 85D, since otherwise, given the same usable capacity, the 265 vs 270 numbers directly reflect the battery-to-wheel efficiency difference?
I don't think any of the supposed reasons make a lot of logical sense. Largely because Telsa published a blog post about the range on the AWD vehicles. If Tesla really did improve charging efficiency and that made up some of the difference I'd expect them to say that much. But they didn't mention it. Same with the usable capacity going down.
So I don't have a pet theory accounted for that difference. I spent a lot of time coming up with reasons during the P85D range debates. Without some good data I'm not going to bother to speculate on what the differences are actually caused by.
I've pondered doing FOIA to ask for all correspondence internally and externally about Tesla's testing of the Model S (old and newer tests both). But haven't sat down and spent the time to write it up. To get all the useful information I'd need to make it fairly broad and I'd probably also need to find a partner more clearly involved in media to avoid paying for the disclosure.
- - - Updated - - -
Breser, What is the real difference between the cars? Aren't you helping lay that bare, yourself? Most people are less concerned with where fault lies in discrepancies, than with what they're gonna get. For 85 / 85D, they can run with 313Wh/mile vs. 316Wh/mile, or 90 MPGe vs. 106 MPGe (Hwy). Pointing that out, a comparison of 70D, to 85D, on a MPGe basis, might raise additional questions.
Speculating, I'd guess that since the highway drive cycle still has accelerations in it, the 85D should do worse than the 70D. It's heavier. OTOH, if both motors are tuned to have less power in the 70D, it may have less ability to remain in single-motor mode on the highway. In my opinion, 70/85 D isn't an efficiency decision. It's a 15 kwh of storage decision.
Of course I'm interested in the actual differences. I haven't tried to explore the reasons for the differences because it's really hard to explore outside of Tesla. Just exploring the difference alone is hard to do with real world tests.
In fairness I don't think the EPA test results are particularly applicable to a lot of people. The average speed over the test cycle is still something like 45 mph (this is based off memory and the exact number might be a little off, but I'm pretty sure it's around 40-50 mph). Yes even the highway drive cycles have that low of an average speed. But they are still useful to compare models.
But stopcrazypp's theory here is that the numbers are distorted by a significant improvement in charging efficiency. Which makes those numbers useless to compare because the Wh/mile and MPGe are wall to wheel.
I think there is obviously some efficiency gains in the 70D. The question though in 70D vs 85D is really about the trade offs between the extra range vs the extra efficiency. Everyone is going to have different answers there because everyone has different use cases.