(Inspired by a random comment I happened to see go past just now, which was drawing one of these inferences and it wasn't entirely clear to me that it was justified in so doing.)
It depends whether someone was seriously saying something like "Adopting DBus is the best thing since we threw out Bonobo" or more jocularly saying "libstartup-notification is the best thing since sex".
I think most people use it casually to mean "Y is even better than X" but logically the only sensible inferences would be "Y is not as good as X, but is the next best thing" or "Y is the best thing in the timeframe starting with X".
X marks the start of the time interval under consideration, and sets an order of magnitude for the level of goodness that Y must have to be interesting. Generally I'd consider that Y would have to beat X for the statement to be interesting, but sometimes they're not really comparable.
Well, yes, but that's the easy bit: clearly it means that there hasn't been a worse storm in the interval between 1984 and now. The hard bit is, what do we infer about what happened in 1984, if we don't already know?
We probably presume that there was some sort of really bad storm in 1984; if there wasn't, you probably wouldn't have said 1984, you'd have said "worst storm since 1927" or some other earlier date. But was the storm in 1984 worse than this one? Perhaps it was (on the same grounds). Or perhaps it wasn't as bad ("this storm was so bad it pushed 1984 off the top spot"). Or perhaps you yourself aren't sure whether the 1984 storm was worse or not, and are making the strongest statement you're reasonably confident of.
Or, conceivably, you might have said 1984 not because it was the date of a storm at all but merely because that's as far back as your records go.
1984 would have to be a worse storm than the recent one, because otherwise the recent storm would be the worst one since a longer ago storm that was worse than the 1984 one. This is the law of weathermen: always compare things to the most historical thing you can.
Wouldn't you then say "worst storm since record began?" Saying 1984 would be too ambiguous - people would probably think that there was a worse storm in 1984, but they wouldn't think that there was a worse storm "when records began"
Well, this is precisely the question under consideration! From the poll results so far, it seems to me as if only about 40% of people would naturally assume there was a worse storm in 1984.
Certainly saying "since records began" or better still "since records began in 1984" would be clearer; the question is, would it be so much clearer that you're entitled to infer that that's not what I meant or else I'd have said it?
For some things I'd say "X was good and Y is as good or better" and for some things I'd say "X is just marking the time period" - for instance "Y is the best thing since sliced bread" would mostly (to me) suggest a time period whereas "this is the best weather since 1990" would suggest (to me) that the weather in 1990 was especially good and the weather now is comparably good, although it is possible it would suggest that the weather now is better than all weather since 1990 but still not actually *good* depending on, er, the weather.
I was briefly confused by this until I realised you were using "strictly" to modify "best" rather than "since".
I do keep wondering if part of the problem is a confusion over whether the "since" is strict, i.e. a confusion between "Y is the best thing in the interval (time of X, now]" and "Y is the best thing in the interval [time of X, now]".
Actually, thinking about this again, I would say it either implies "X was good and Y is nearly as good" or "X was good and Y is good too, but no implication is intended as to which is actually better". I am, in fact torn between these two. If you say that [X flavour] crisps are the best things since pizza, I think that probably implies that pizza is better. Or else surely you'd go further back and say that [X crisps] are the best thing since roast beef or something! :P
I've been trying to think of a particularly good example of that usage. Best I've come up with so far is that one might just about plausibly describe something as "X is the biggest step towards international harmony since World War II", intending to imply not that WW2 was itself a step towards harmony but that it wiped out the effect of all previous such steps, making X the biggest step whose effects we're still feeling.
"X is the biggest step towards international harmony since World War II"
I first interpreted that as "The X Window System is the biggest step towards international harmony since World War II". Although you might even say that was true, if sufficiently cynical and/or facetious.
I think it just means Y is pretty good. A literal interpretation might be "This is the greatest achievement since X happened" but nobody really means that. If you say "The best thing since sliced bread", it's pretty clear that you're not literally claiming to have examined the history of human achievement since 1928 (when Wikipedia claims sliced bread was invented) and found that you've been able to make an objective assessment that Y is better than hydroelectric power, The Dark Side Of The Moon, the universal declaration of human rights, or hotpants. It just means that it is a comparatively sit-up-and-take-notice, changes-the-way-we-live innovation. Or isn't, but that's the kind of hyperbole that you felt like invoking.
If someone actually said "the best thing since sliced bread", then yes, it's almost certainly pure hyperbole and if they meant it literally they have only themselves to blame when people don't realise that. But if they said something more specific, such as calling someone the best actress since <some actress who is actually within living memory>, it's not so clear. There's probably a general hint of "comparable to" or "merits mentioning in the same sentence as" in there, at least.