Whether you want to speak a dead language or not is your preference and I see no reason not to do so if you
can. But one thing that I hope you understand is that sometimes, we just don't understand the language very well. Biblical Hebrew, with its entire Biblical corpus, is still too meagre to exhaust everything we want to know to use the language for
everything.
You want to converse in Sumerian? What would be your model for conversation? You're not going to build a conversation based on royal inscriptions or economic texts, which comprise the majority of the texts, nor from highly stylized poetry. There are surviving philosophical dialogues, but this too is far too idealized and stylized. (Not to mention being few in number) If nobody left us records of a conversational language, then it's just obvious that we can't reconstruct
that aspect of the language. Even the languages with extensive literature like Latin have the same problem.
Not only was Sumerian written for thousands of years after it no longer the vernacular, but what was written has preserved only a part of the language. The surviving texts consiist primarily of highly conventionalized administrative documents, academic word lists, and poetic compositions; there is very little literary prose. As a result, one mus talways keep in mind that we are dealing with highly formalized forms of verbal art far removed from any putative language of the streets, constrained by certain conventions with restricted rhetorical scope.
Piotr Michalowski, "Sumerian", in The Ancient Languages of Mesopotamia, Egypt, and Aksum
But let's say we ignore all the pragmatics of the language. It's the modern incarnation, so the language's style can certainly change. Let's call it language evolution and we could speak neo-Latin, Modern Hebrew, or whatever....assuming that we at least understand the basic structure of the language fairly well. The problem with Sumerian is that we don't. It's not just "some grammar", it's a lot of the grammar. Much of the verbal system remains a mystery. From the same book, let me just cite a few sentences.
Phonology: "The phonology of the language is not well understood, and it is fair to say that it will never be fully recoverred."
Pronouns: "Nothing is known about inanimate third person, although it is possible that this function was fulfilled by
ur5 (or
ur5-bi). As already noted, not all forms are attested."
Adjectives: "No proper study of adjectives exists; recent grammars contain limited information on this category."
Valence: "Matters of valence in Sumerian have been disputed, but no concensus has been reached."
Tense and aspect: "Opinion is divided on whether the two forms of the Sumerian verb differ in tense or in aspect. ... There has been much discussion of the exact meaning of these words [hamtu and maru] as well as of whether these technical terms describe the Sumerian verbal forms or their Akkadian translations."
Mood: "The traditional description of modes distinguishes between pairs of homophonous prefixes that differ in meaning depending on the mood. ... There are reasons to reject this interpretation; certain modal prefixes are indeed usually associated with one aspect or the other, but this results from the semantics of the mode and not from any formal constraints."
Conjugation prefixes: "The prefixes that fall in this position constitute the most controversial part of Sumerian grammar. No two Sumerologists appear to agree fully on their form, meaning, etymology, and identity; the number of ranks that they occupy is equally disputed."
Syntax: "The syntax of Sumerian is perhaps the most neglected part of the grammar."
I could go on more, because almost every aspect of Sumerian grammar is disputable. I know of Edzard's grammar which you mentioned, and I have read quite a bit of it myself, and while I do think it's a great work, it's not conclusive. (The chapter which I cited, "Sumerian", in
The Ancient Languages of Mesopotamia, Egypt, and Aksum, was written after Edzard.)
So, this is my take on it. If you want to speak Latin, ignoring ancient pragmatics because now it's spoken in modern settings, go for it. No problem, you could still speak grammatically correct Latin even if your phrasings might look awkward to the Romans, but they're dead and you're not. Modern Hebrew already became a new language on its own, and people distinguish the two for a good reason. If you want to speak Gothic, it's going to have to be extremely cautious project because, as far as I know, we still don't fully understand the aspectual (imperfective-perfective) distinctions of Gothic, which is something you really need to understand to speak a language that makes such a disttinction.
If you want to speak Sumerian, I'd say it's damn impossible unless you want to re-create a Sumerian-based conlang from scratch because as you saw above, almost every aspect of Sumerian will need to be reconstructed. So, my bottom line is this: some ancient langauges cannot yet be revived for a good reason, they're still in the stage of decipherment, and we don't understand them as fully as we'd like to.