It's mildly worse than that, because the US uses metric side-by-side with US customary units, which at least for volume use the same unit names as Imperial but are smaller.
(Does the US use Imperial as well? My most common experience with US vs imperial measure is in pints of beer, although Canada has caught on to the 16 oz pint, and don't even start me on 12 oz "pints".)
The US does not use Imperial measurements that I know of, which can be confusing to Americans if UK mpgs are interpreted as equivalent to US values.
But there's more than one kind of ounce in the US. Fluid ounces are volume whereas dry ounces are weight/mass. From memory - and I memorized this from cooking - 8 fluid ounces (one cup or a half pint) is 8.337 dry ounces (assuming the approximate density of water).
Then there's metric horsepower vs. other types. It's clearly so much more natural for a horsepower to have a German name (pferdestärke!) and mean 735.5 watts instead of 745.7 watts (numbers copied and pasted, not checked).
> From memory - and I memorized this from cooking - 8 fluid ounces (one cup or a half pint) is 8.337 dry ounces (assuming the approximate density of water).
Why memorise this to such precision if you're already approximating away the density? The job is already done for you, 8 dry ounces is approximately 8 fluid ounces, whether its water, milk, oil, or orange juice. If you need to measure liquid mercury, you're not working in the kitchen; if you need greater precision, you need to measure mass or volume directly as appropriate.
Because that was the precision used by the particular author I got it from. It's not that the fourth s.f. really matters. I think the 7 actually helps my memory by making the number distinctive and not just 8 1/3.
It's like remembering that body temperature is 98.6 F. It's obviously false precision and it's not even an accurate average, but it's the way it's traditionally recalled - plus it works out to exactly 37 C.
>8 dry ounces is approximately 8 fluid ounces
I did write "cooking" originally and it is a valid point that when cooking as opposed to baking usually a little more or less liquid is immaterial.
But I meant to include baking and, say, mixing drinks.
A 4-5% difference isn't always important, but it's often noticeable. I've been using a digital kitchen scale for >20 years, and I just find volumetric measuring cups annoying compared to going by weight. I don't want to go back and forth filling a container and checking it on a level surface, to eyeball a meniscus, to wait for every drop of a viscous liquid to fall, or to sift flour.
I don't cook or bake with liquid mercury, but I use sugar, which is ~7 ounces per cup, and flour, which is (according to me) 4.25 to 4.5 ounces per cup. That's considerably different from 8 = 8.
(Does the US use Imperial as well? My most common experience with US vs imperial measure is in pints of beer, although Canada has caught on to the 16 oz pint, and don't even start me on 12 oz "pints".)