For its time, I think Fahrenheit was probably quite logical. The "0" point was that of a freezing mixture, and the nominal "100" was human blood temperature, albeit a little in error unless the subject had a fever.
I was going to suggest it was a centigrade earlier, when someone referred to Celsius that way. However it was 96 as it was easier to make a scale by dividing stuff into equal halves multiple times (32 below ice melting/water freezing, 64 above it) than making tenths. The freezing mixture (or the coldest temperature one winter in Danzig where Herr Fahrenheit lived) at 0, 96 as blood temperature, and it just happened to work that water freezing was a third of the way along.
As for "for its time", Celsius was in its present form in 1744 (the original 1742 scale that Celsius himself made had 0 as boiling and 100 as freezing - influenced by Delisle's 1732 scheme but using 100 for freezing, not Delisle's 150. Christin's 1743 scheme, independently came up with, had it the other way around. Linnaeus flipped Celsius' scale (independently of Christin) when Celsius died in 1744). Réaumur had 0-80 with the same anchor points in 1730. Fahrenheit was 1724 - not much earlier.
Rømer, whom Fahrenheit was influenced by, used the boiling point of water and freezing point of brine (so water froze at 7.5 and boiled at 60). Newton's slightly earlier scale had the freezing point of water (0 degrees) and body temperature (12 degrees) as scale definers. Effectively, Fahrenheit had two existing scales to work off of - one where brine freezing (roughly - it was recalabrated so that water freezing was exactly 7.5) was the zero point with water boiling the top anchor (at 60), and the other where water freezing was zero and body temperature the top anchor. He went with a mix - and chose the brine and body, rather than the water freezing and boiling!
Of course, most of the metric system is based off a billionth (10^-9) of the distance between the North Pole and Equator through Paris (the centimetre), and that (deliberately) rather arbitrary and unuseful (it took them a while to calculate the quarter-circumference properly) definition of a centimetre - at least some guy's foot can be roughly visualised quite handily! - is much more random and obscure than body temperature or the freezing point of a sea-esque brine solution.
In the UK we're far less likely to come up against zero degrees F. and a hundred degrees F., but these extremes are far more commonplace across the US and I can understand why it makes perfect sense there.
Even with our weather, Fahrenheit makes sense - in fact more so as 0 and 100 are rare outliers (if not quite actual extremes) here and so everything fits like its a "percentage hot" scheme. There's good reason why we picked it up and carried on using it long after most places had decided on Celsius (Réaumur was popular in France and Russia, but fell in favour of Celsius in the 19th century).
IME people use either celsius or fahrenheit consistently, I've never known anyone who switches between the two depending on whether it's hot or cold.
I guess you weren't in the UK in the 80s or 90s.
The topic of this thread is ""Pheww it's 90 today" or "Pheww it's 32 today"?" The lukewarm-sound of 32 is the reason we switched to Celsius - the media needed to more extremely convey freezing weather to sensationalise it and so in winter we started using Celsius in the same way that tabloids use Fahrenheit today. And it took time to do so fully because of days like today - 90 conveys the heat better than 32, just as 0 conveys the cold better than 32!
We, as a nation, switched organically from pretty much entirely working in Fahrenheit and not understanding Celsius, through pretty much being able to cope with either but preferring Celsius in winter and Fahrenheit in summer, to pretty much entirely working in Celsius and not understanding Fahrenheit (I can just about do it - and my parents, who taught me Fahrenheit, have fully lost it).
Maybe we should use Kelvin, that's the logical one where 0 is actually 0 and not an arbitrary midpoint!
Rankine has that too - only with the finer gradient of Fahrenheit's scale. Both named after Scots who must have worked together in Glasgow (Kelvin made his in 1848, Rankine his in 1859) :P