I think natural language semantics is a bit of a fool’s errand, at least as it is conceived and practiced throughout most of the world’s linguistics and philosophy departments. One encounters a myriad of little proofs of this daily if one keeps one’s eyes open.
I’m pretty certain that by ‘squeezy honey’ the manufacturer intends ‘honey in a squeezy bottle’. I am reasonably certain that by ‘pure honey’ and ‘clear honey’ they do not mean ‘honey in a pure bottle’ and ‘honey in a clear bottle’ respectively, but honey that is pure and clear. They don’t mean colourless by ‘clear’ but simply not turbid or cloudy.
No one would have any difficulty understanding the phrase ‘squeezy pure clear honey’ appearing as it does, in context, on the label of the bottle to mean pure clear honey in a squeezy bottle. Everything is just as it says on the tin. But isn’t it interesting how the adjectives seem to modify the noun ever so slightly differently? It isn’t the honey that is squeezy but the bottle of honey although it is the honey which is pure and clear.
It isn’t true in general that ‘squeezy X’ means ‘X in a squeezy bottle’, because that doesn’t work for X = ‘bottle’ (or ‘ball’ or ‘cushion’ or ‘toy’ or various other things).
We seem here to be dealing with a case where the precise meaning of ‘squeezy pure clear honey’ is fixed by context in which it is used. The mere fact that it is written on the label guides our interpretation. If we had encountered the phrase out of context, it might not have made any sense to us. This sort of thing, I think, makes it hard to credit the prevailing theories of meaning current in philosophy and linguistics departments that seek to associate general and context independent rules with linguistic expressions.