Both.

With the exception that there are no best guesses. Only guesses.

:up: And, consider how English is pronounced and used around the world. I have Greek textbooks that inform me that zeta is to be pronounced dz, zd, z - I make my choice (dz). Yet there is a right way and a wrong way - which in an interesting way extends an invitation to look more closely at just exactly what "right way" and "wrong way" mean and under what circumstances.

The hieroglyphs can only be interpreted and not pronounced, but thats alright because nobody uses it as common language now anyway. So they need to be translated by experts to have any use. Pronounciation is redundant.

As you’re no doubt aware the recordings of the Egyptian language possessed no vowels. So it is guesswork, but not blind guesswork as they base their guesses on Coptic dialects and reiterations in other languages at the time.

It’s been some years since I was interested in this subject but I recall they used mostly derivations of Coptic dialects rather than sourcing Ancient Greek texts and such?

The hieroglyphs can only be interpreted and not pronounced

Well, actually, many of them can be pronounced. They're mostly consonant signs, so they're like letters and letter combinations. Save for the class of signs we call determinatives, those are true ideograms.
As Arthur Rupel said, we're unfortunately missing most of the vowels. We have and alef, an "y"-sound and a w thats similar to the Arabic waw. But there are more vowels in the words, and we don't know what they are and where exactly they go, because they're not written out. It's actually pretty similar to written Hebrew and Arabic.

Greek inscriptions helped to translate the language since there are a few biliteral inscriptions - the Rosetta Stone being the most famous.
Coptic on the other hand is a later form of the Egyptian language. Coptic has incorporated a lot of Greek vocabulary and uses a different alphabet (including vowels, thank the gods), but the grammar and a lot of words are similar. So that helps with reconstructing both the meaning and pronunciation. Still, there's a 2000 year gap between classic Middle Egyptian and Sahidic Coptic. If you consider how much the English language changed in the last few hundred years you can see how it will offer clues, but no certainties, in reconstructing the ancient language.

So yeah, the answer is "both". There are clues and we can reconstruct an approximation, but it's still a lot of guesswork.
Egyptologists don't really try with the correct pronunciation most of the time. They just use the existing consonants and stick random vowels in between for convenience.