The power consumption of the Semi is mostly the load of the actual work being done to overcome frictional forces, so there's minimal opportunity for improvement beyond Tesla's stated specs in 2017. The range of uncertainty is only like 5-10%.
I've calculated an estimated 1.1 kWh per mile of air drag power consumption at 75 mph using Tesla's 0.36 stated drag coefficient and 0.8 kWh per mile from rolling resistance.
Tesla can only do so much to make a big box with a 10 square meter cross-sectional area actually aerodynamic. Most of the design win was already achieved in 2017's design simply by making the front of the truck smooth, sloped and bullet-shaped.
It's also extremely unlikely that Tesla has found a way to revolutionize rolling resistance, and meanwhile their 82k lb fully loaded weight is more than the 80k lb legal limit for Class 8 diesel trucks, which increases the rolling resistance by 82/80-1 = 2.5%, all else being equal. The energy loss from rolling is due to vibrations, the material properties of rubber, and the road surface. Tesla may have squeezed out some tiny gain in this area but there's not much room for improvement.
The only way Tesla could get a major improvement in these areas is with platooning, such that one truck lowers its wind load by drafting behind the truck in front of it.
In the longer run, the only possibility for terrestrial freight transport at these speeds to require significantly lower loads is to do the platooning in a tube that contains the airflow and has either rails or smooth, clean pavement. One of these from Boring Co could do the job:
View attachment 862272