Are you the only subscriber connected to the transformer, or is it shared with other homes (like your own)? Even a brand new dedicated drop and meter will not be immune from misbehaving loads connected to the same transformer. The transformer provides some buffering against primary-side problems, but even in some cases the power quality from the power company can be an issue. Professional electricians and most power company first-level techs won't be able to identify this because they're using tools that only reflect the power quality over a period of time. You won't see it as a "voltage sag" -- it will be an instantaneous surge or drop for fractions of a second.
In my case, a bad lighting ballast in a shed caused this at a property I own. If you placed a volt-meter on the service you'd see it's stable at 240V, because the meter is just measuring RMS power over a period of time. I found my problem by chance - I noticed it happened when I turned the lights on in the shed. Other people have found that it occurs more often in summer vs. winter (go looking for A/C units), etc.
...and I doubt that your issue is 208V vs. 240V in the service center. As Jerry said, see if it happens at a local RV park or another Tesla owner's home.
I live in a typical Bay Area suburban neighborhood, so I'm definitely not the only home on that transformer. I know it's possible that there could be some transient quality issues with the incoming supply that won't show up with a multi-meter. But, Tesla's logs also don't tell them what caused the charge rate reduction. You would think if it's something like an incoming voltage spike, that would be logged, right? If the on-board computer software can react to it by reducing charge rate, then it must be able to detect it and log it. I guess the only way to be sure is to put an oscilloscope on the line to trigger on glitches or leave it on infinite persistence. It's just a hassle to do this with my wife and young kids around the garage all the time.
I really suspect the car itself is more likely the culprit. And in that case, 208V and 240V definitely matters. If the on-board charger, or cabling, or contactor or something is marginal, then 10kW vs 8kW could make a difference. The amount of power passing through those components will be 15-20% different. One clue is that Tesla actually checked the logs and claimed the problem was happening due to my incoming voltage being too high. At the time of occurrence, they logged 246V and 39.49842A. This is 9.72kW. In contrast, when they charge the car at the service center, it's only about 204-205V and 40A, which is barely above 8kW. But they only test it for one night at a time, which is not conclusive because it could take a couple of weeks or more for this to occur.
They dropped that line of reasoning when I made it clear that 246V is acceptable according to PG&E standards (
Link) which specify service tolerance to be +/-5%. For a 240V service, that's 228-252V. That's when they agreed to replace the on-board charger last year. It seems, to me that the frequency of occurrence has reduced, but not by much. At the end of the day, it's an occasional annoyance, it's been almost 2.5 years, and I've already given-up since last year. I just accepted that this might be normal for a lot of people since the 5.8.4 FW update.
