few notes:<p>- "Fuel consumption of 9.4 L / 100 km" is not "the right form" is a modern take, a more classic and still valid SI one is km/l which is very similar to "Fuel consumption of 9.4 LPK";<p>- "25 {kW/h,kW⋅h} to boil a tank of water" are both strange forms, kWh is the common accepted way to express energy, no center dot multiplication needed;<p>- multiple prefix (vs power-of-10) and bare prefix are as well commonly accepted, I see no reasons to consider them wrong, while I can state formally wrong "2 kilograms of rice" because we do not measure the mass but the weight so it should be 1.961daN where deca-newtons are commonly used because 1 daN is roughly 1kgf commonly shortened to 1kg as we can commoly count 1kg[f] == 1 daN... For instance climbing equipment in the EU use daN to express maximum loads of connectors, ropes etc because of that;<p>- multiple quantities much depend on industry and conciseness, as we do not write units in table values but only in headers we tend not to write them three times in a row where from the context is clear what numbers means.<p>The biggest issue is makes habits changes. Actually we should not use km/h as well, since for SI base unites are m/s, but 3.6 is not an easy conversion like kgf/daN, so in the EU we keep using km/h, something meaningful in the past, when we go by horses and feet, but not much needed today.<p>Not to count software, where often recognize "°C" (two chars) BUT not ℃ (U+2103) and so on.