This isn't really an "either-or" thing. Deciphering a language is a very, very long and slow process. There's no single moment you can point to as "the moment" it was deciphered.

In the world of deciphering long-lost scripts, though, there are rare "eureka" moments; but even there, those are the rare exceptions, like the Egyptian hieroglyphic and Linear B scripts. It's just that the rare exceptions are very well-known precisely because they're so unusual and impressive.

But with languages, that's never going to happen. The Hittite language began to be deciphered in 1915, but wasn't in a state that you could call "fully understood" until 1950 or so. The decipherment of Luvian began in the 1920s, and it still isn't "fully" understood. Etruscan is in a similar state, though that's even harder because it isn't an Indo-European language. (Take my uses of "fully understood" as approximate: of course there's always going to be more to learn about a language, so there's a bit of arbitrariness about what you can call "fully" understood.)

Generally when a dead language isn't fully understood, it's not because of difficulties in the decipherment process (because humans are clever), but mainly because of a shortage of source material. You need an awful lot of context to decipher lost languages. The more context you have, and the bigger the corpus of texts you have, the closer you are to a minor "eureka" moment: in the case of Hittite, a large text archive was discovered at Hattusa in 1905, and the decipherment began in 1915; the more material turned up, the more the language was understood. In that sense it's not a slow process. But it's in the nature of things that texts in lost languages do not turn up at a terrific pace. That's the thing that determines how close you're going to be to understanding something new about that language, so unless a large corpus of texts has turned up very recently, there isn't really any sense in which any language is "closer" than another to being deciphered.