It's been clear since the 1990s at least that Julius was certainly a spy. Nobody contends otherwise anymore, though it used to be something an item of faith among some far-left scholars that the whole thing was a set-up of some sort (as improbable as that might seem in retrospect).

Whether that guilt of Julius justifies the punishment (execution) is a more contentious question. The status of Ethel remains contentious — David Greenglass later admitted he perjured himself on the subject of Ethel in order to get his own wife (Ruth) off the hook, and the VENONA decrypts in various ways indicate Ethel was not any kind of major player (she is referred to as ETHEL and not some kind of code-name which is a sure sign she was not some kind of actual agent). Now did she know what Julius and her brother and sister-in-law were up to? It somewhat beggars the mind to suggest not. Was she complicit? Probably. Did she deserve to be executed for that complicity? I think different people can come to different conclusions on that.

As for whether one should give them credit for the Soviet nuclear program — no, not really. The intelligence that the Rosenberg ring gave on the atomic bomb was very meager. At best, it was useful in determining that the better information, given by people like Fuchs and Hall, was not fabricated. Greenglass was not a very technical person and the kind of things he passed on were pretty sketchy and basic — less information than you could get from an encyclopedia today. It was not blueprints or key numbers, it was very basic stuff (on par with "they are researching implosion" and "here are some of the basic experimental setups they are doing" and "here are a few other ideas they are playing with").

There is no serious evidence that the Rosenberg data was all that useful to the Soviets. They did go over the Fuchs data very closely but it is also clear that they did not trust any of their intelligence data entirely and Beria required total re-confirmation of everything (and did not share it very widely). It is not clear that the spy data on the whole actually aided the Soviets all that much in their acquisition of the atomic bomb. It certainly wouldn't have been required for them to get a bomb in general — they definitely could have made a bomb without it, it is just a question of how much longer it would have taken. The most optimistic estimates by historians of science and technology who have looked at this suggest that the Fuchs data might have saved them 6-12 months. I suspect it is much lower than that. The major thing that set the timetable of the Soviet bomb project was not scientific/technical knowledge of the sort spies have, but access to raw uranium resources and the industrial processing necessary to turn it into fuel. Intelligence data was not very useful for either of those (Fuchs was a co-inventor of the gaseous diffusion process, but otherwise none of the Los Alamos spies were connected to that kind of work at all).

This doesn't get the Rosenbergs "off the hook" in any meaningful way — spying is spying, whether the enemy uses the information efficiently or not, whether the information was that useful to them or not. Of course, at the time of the spying by Rosenberg/Greenglass, the Soviet Union was an ally and not an enemy, a factor that meant nothing in American courts but was key to Fuchs getting a relatively light sentence in the UK (even though he continued to spy well after the war ended and the Cold War had begun). Again, I leave it to others to debate the justice of the Rosenberg verdict and execution, but in the specific cases of were they guilty (Julius definitely, Ethel maybe) and did they help the Soviets much (no), the factual answers are more or less clear.

On the use (and non-use) of the intelligence data in the Soviet program, see Michael Gordin, Red Cloud at Dawn. On the Rosenbergs specifically, I am a fan of Ronald Radosh's book, which seems generally fair.