While the statement in question is a strawman argument, it is certainly not wrong, as general as it is. Ironically, explaining why will also answer your general question.
- It is a strawman, because you have not told us who said it. You posit that the statement is correct, and you want to know if that's true and you want to know details. That's a usual way to phrase a question, there's not much wrong with that. Except: It shows that you have no own opinion, did not conduct any prior research on your own. For benefit of the doubt, we might assume that you have an unexplicable, yet obvious reason to believe the statement true. This depends on philosophical definitions of truth.
And guess what, that depends on syntax. Can a question be logically wrong. Is a question formed through syntax, or through punctuation. And is there even a difference if particles serve for punctuatuation hmm. q.e.d.
Syntax is essential if you want to be a good translator.
In essence, [semantics underspecifies], and syntax is needed to constrain the specification. Point in case, Syntax is essential is a very general statement, so general that it is devoid of meaning and thus to be rejected. The added constrained in the conditional clause if you want to be a good translator doesn't change that. Why? There are two ways to look at it:
you are really asking whether you want to be a translator. Nobody can tell you. Also, the correspondants so far don't want to become one, and thus reject the proposal. That is not very tactful, and this reading is difficult to justify from syntax alone (basically I presuppose an inadvertant freudian slip; full disclosure: I'm randomly reading if as its German cognate ob "whether" as a matter of habbit, trying to understand how these two different senses could diverge, which is of course also a matter of syntax, insofar ob can also mean "because of" in different syntactic environments, primarily marked by Dative case).
We might even suppose an aequivalence relation, so that an invariant reading of the statement would be you want to be a good translator if syntax is essential (if you want to be a good translator. At least, this is left open by the question. Basically, one might wonder whether the skills of a translator have merrit for more essential ...
More naturally, we read a questionable statement saying that syntax is prerequisite for a good translation.
a) that's evidently true, if a text without syntax cannot be.
b) I said it's underspecified, and this would be constrained through syntax. Well, the given condition constrains only the subject, but not the instrument or quantity. It leaves one guessing "important for what", "how important?" or just "how?". And in my naive understanding of syntax, the statement leaves room for those constraints, it leaves them open:
syntax is essential for ... if you want to be a good translator
On the one hand, this can be understood as an implicit question. On the other hand, there are simple semes that are themselves underspecified and can thus easily fill the void, i.e. to answer the potential question in an evasive manner. Point in case, it would be utterly redundant to say "for you, if you ..."; And it would be likewise highly underspecified what "for text-understanding" means--which leads to my conclusion:
- There is no generally specific answer. The question is in that sense too broad. Challenges of translation have to be decided on a case per case basis.
But it also connects to my first bullet point:
Nor do I have professional experience in translation. I suppose trivial translations use an intermediate step through internal representation, so to speak. That however brings up two prominent problems of linguistics: Universal Grammar, insofar I cannot explain myself "internal representation"; The Sapir-Whorff hypothesis, insofar some concepts translate badly between languages which also includes idioms.
So finally, it is at least important to recognize what is not a transparent syntactic compound, but a lexicalized pattern. Because it is important that translations sound natural. In that sense, translation is more of a craft than a science, with its own idioms of translation practice that is individual to each language culture.