In the first case you deduce the time. And, until the law of gravity is established, in the second case you induce the arc.
In the first case it is entirely reasonable to be perform your little verification step because man-made objects sometimes fail. In the second case there is no need to second guess the universe because your intuition that the laws which govern things like the flight of a ball are not susceptible to revision is correct.
I don't think you're butchering anything, I just think you have to be able to recognise when you are using the inductive method and when you are using the deductive method.
Next question would be: do these two methods exhaust the range of methods we employ to acquire knowledge? If not, what are the other methods?
I'm going to butcher stuff, but could someone please do an ELI 5 why you can't just infer that the clock isn't working? When I look at clocks I look at the hour hand and minute hand. Then I verify my believe by looking at the second hand. That process takes longer than a second, so I immediately notice the clock isn't working and that I cannot be sure whether it's noon.
I actually have a bit of a tick when watching clocks in that I repeat this process multiple times at the weirdest moments when there is a clock available. I immediately notice a stopping or non-working clock.
I always felt like that -- in first hand experience type of stuff -- we know something is true because we've seen it a trillion times before, like gravity and throwing a tennis ball.