I think people in the U.S. would adapt if we started using the metric system more frequently. We already use many of the prefixes in the computer age anyway: kilohertz, gigabytes, and so forth. It is merely a matter of what one grows accustomed to. Inches are no more and no less intuitive than centimeters as a size referent, though they are more difficult to calculate and convert in your head.
As far as temperature goes … I don’t know if I agree that either system is inherently superior. In both systems there’s a great deal of headroom between “it’s very hot out” and “water boils” that we simply don’t use much. When we’re speaking of the difference between 50C and 80C, or its equivalent in Fahrenheit, it’s just numbers on a dial to most of us. We don’t have a direct sensory connection to the numbers.
As a civil/environmental engineer who used to teach chemistry and physics, I think I have a good perspective on the unit situation in U.S.
Chemistry and physics are taught universally in U.S. schools using the SI system.
In civil engineering, survey distances are made using feet and decimal fractions of feet (e.g. 150.58 feet). Material volumes are measured in cubic feet or cubic yards. Liquid volumes are measured in gallons.
In environmental engineering, for measuring concentrations of environmental pollutants, we use mg/kg or mg/L.
Ambient temperatures are measured using °F.
For hydraulic and fluid dynamics problems, I prefer SI to avoid issues with lb[sub]m[/sub] vs. lb[sub]f[/sub]. Many references provide data in non-SI units, though, so I often find myself doing a lot of unit conversions.
In short, we have a complete hodge-podge, and U.S. engineers have to know both systems.
By the way, the National Council of Examiners for Engineering and Surveying (NCEES) refers to the system of units used in the U.S. as the U.S. Customary System (USCS).