My straight line is at approximately 156 Wh/km.
What should it be?
That's correct for an AWD.
EDIT. This is surprising:
156 Wh/km * 500kms = 78kWh
Isn't the battery supposed to be 75kWh and 74kWh usable capacity?
No. There are CAN bus readbacks I've seen showing 77-78kWh (including the buffer) for a brand new car. In the EPA test they
discharged 79.2kWh https://iaspub.epa.gov/otaqpub/display_file.jsp?docid=46585&flag=1 from the battery (but keep in mind these
may not be the same kWh - there is no reason the BMS needs to read true kWh, and it would have to be calibrated to match the calibrated equipment in Tesla's EPA measuring lab, anyway).
Furthermore, you'll never actually be able to draw the full 77-78kWh, because that would require driving until the battery is completely dead and the car stops moving, and the last 20km of draw would need to be conducted at very low consumption levels so as not to droop the battery into automatic shutdown.
Finally, while ultimately 156Wh/rkm*500rkm = 78kWh is probably no coincidence, it is a bit misleading. And again, 156Wh/rkm is NOT what is used for that (not too useful) Energy Consumption screen. (153Wh/rkm is used)
Specifically it is misleading because each rated km you use will show up on the trip meter as no greater than ~146Wh/rkm. So if you travel far enough to use 100 rated km
on the battery gauge, you'll see no more than 14.6kWh on the trip meter. And it may be as low 14.2kWh. This means you have at
most 73kWh (500 rated km - assuming you can get to that * 146Wh/km available for use as shown on the trip meter, before you get to 0 rated kilometers (and from there - if you get there - you would have to rely on using the buffer at a very low output power).
It's also misleading because at high SoC, the projections turn out to imply that you must fully use the buffer to get the predicted range. Whereas at very low SoC, the projected range, you will find, appears to show projections assuming nearly zero entry into the buffer. This behavior is just a function of the 153Wh/rkm constant being used for the projection, whereas actually you can do no better than 146Wh/rkm. Easy to test this claim out yourself. In some ways this behavior is good, but it does cause confusion.
You can check out my current summary/tracking here (which is subject to change if errors are pointed out - I did play fast and loose with Wh and kWh, so be careful about that):
Model 3 Lines & Constants Pre 36.1