that is simply not true - there is a big video GCN did on this.
I looked up the video, if you are referring to the below video, it actually supports the argument of
@comanchepilot (avoiding the absolute maximum and minimum usable):
First of all, the type of tire/wheel he is using is fairly specific (may not be applicable to other road bikes) and he is testing on rough roads. 28mm tubeless tires on Zipp hookless 303 Firecrest Wheels which are 25mm wide. He is 72 kg, and he is doing the test on chipseal roads which are rougher than typical asphalt or concrete roads (conditions were also damp, although not fully wet).
Chipseal - Wikipedia
Even so, the results he got was at 62.5 psi he was 0.3-0.4 kph faster than at 72.5 psi (the max allowed on his tires). He then lowered it to pressures recommended by the manufacturer's tire pressure guide for his conditions (53.7 psi front, 57.1 psi back) and only gained 0.1 kph (which he mentioned was probably not statistically significant). He then lowered it to 42.5 psi and actually went slower than at the pressure guide based pressures (although not slower than at 72.5 psi). So basically he only gained something significant when he went below the absolute maximum allowed by the tires, otherwise it was either not statistically significant or he actually went slower when he lowered pressures below manufacturer recommended.
In the end, he referenced the document here, which is mainly based on rough roads (poor-condition asphalt or dirt roads). Basically the idea is that the power required from the rider is both from rolling resistance and vibration losses (which play a bigger role on rough roads). This graph (which he showed in the video also) basically illustrates the idea:
https://www.sram.com/globalassets/p...emeffeciency/pdf-downloads/tse-explained2.pdf
As you can see, rolling resistance goes down with tire pressure increases. Vibration loss goes up. There is a sweet spot that optimizes both. That sweet spot pushes downwards on rough roads, and upwards on smoother roads. So basically due to the fact that a road bike has limited suspension to absorb vibrations, loss from that can be higher on rough roads. This may not be applicable to cars where the suspension system that supports most of the weight of the car can do a lot more absorption of road vibrations.
If you apply those results to this thread on Tesla tires (which as above may not be applicable in the first place due to major differences in the suspension design), all that tells you is to avoid pumping your tires up to the max allowed by the tire (50 psi or 51 psi) to reduce vibration losses on rough roads. It doesn't say necessarily that reducing pressures further lower than manufacturer recommended will help improve efficiency (instead it resulted in the opposite, going lower than manufacturer recommended was less efficient), especially given rolling resistance goes up as tire pressure goes down.