I ate yesterday, simple past.
I have eaten many times--cannot use this tense with specific time. Many times is not specific.
I had eaten yesterday, when my friend appeared at the door. Past perfect for a completed action in the past, when some other action occurred.
They haven't written yet. Time is not specific, and one uses present perfect.
I would leave it to my British colleagues to discuss biscuits. We eat them here, but the term is not used often.
They ate two biscuits two minutes ago. Simple past with specific time.
As Lambie says, drinks are either carbonated or non-carbonated. I believe these are universal terms used in government or official communication.
In the US:
Carbonated soft drinks are collectively referred to as soda, pop, and in some parts of the country Coke (even for carbonated drinks that are not Coca-Cola). Non-carbonated drinks are referred to by name (fruit punch, lemonade, iced tea, etc.)
Regular water can be either bottled or tap (meaning from the faucet). Carbonated water can be called soda water, and still may be referred to that way when ordering mixed drinks, (e.g. a scotch and soda). These days, however, bottled carbonated water is usually sparkling water, or colloquially bubbly water. Fizzy water also works.
In some fancier restaurants, if you ask for bottled water you may need to specify whether you want sparkling or flat.
Recently there are some naturally carbonated drinks such as kombucha which would not be grouped in with soda, as that usually refers to sweet carbonated drinks like Coke. Because it doesn't really fit into any category, just call it by name, kombucha.
Side note: Historically "soft" drinks were those without alcohol. At a large social gathering, for example, there may be a "soft" punch for the children and adults who didn't drink, and a "hard" punch for the rest. These days when you say "soft drinks" people mostly think of soda, but, technically, it does include any flavored non-alcoholic beverage.
Best Answer
My gut feel is that the mightn't contraction is more common in BrE than AmE (but I've no reason to think most Americans find it at all "unusual").
I'm not sure if this is a valid way of using Google Books estimated results, but note these figures...
According to #1, there are three times as many instances of US color as there are UK colour. Since there's no reason to suppose Americans use the word any more often than Brits, I assume the relevant US corpus is about three times bigger than the US one.
All things being equal, we could expect the #2 AmE count to be about 18M, not 12M. But note that overall prevalance of might has fallen by a third in the last couple of centuries. Older texts in the Google Books corpus are more likely to be BrE, so the AmE aversion to might is weaker than the bare figures suggest.
But after allowing for the implication of the above (Americans are perhaps 20-30% less likely than Brits to use the word might in any given utterance), we should still expect the #3 AmE count to be 11K, not 4K.
Regardless of whether the absolute values of those Google Books estimates are accurate, I see no reason to suppose that the ratios aren't meaningful. So I think this validates my gut feel.
The figures for couldn't + color:2.8M and couldn't + colour:0.85M suggest that the US aversion to mightn't doesn't extend to couldn't.