There are specific reasons the OP's examples worked so well for their specific use of visual effects. Jurassic Park, the T-Rex was actually motion-captured as a stop-motion puppet, and later rendered and lit, and then added to the shots in post.
Anyway, I see some benefits to model work. You have a tactile object with literal detailing, imperfections, and character; that translates subconsciously to the viewer that's extremely hard to replicate in a computer. The lighting, while it may be irritatingly difficult to match, is real lighting. However, there are significant drawbacks. The OP posted the Battle of Endor. That entire battle was assembled on the film itself - the physical material - one ship at a time. Each starship was built, filmed, and added on an individual basis. So, while the models may have an edge in reality and may be cost effective, it's also very, very time consuming. The only reason we were able to go from a few X-Wings in a single shot above the First Death Star to dozens of starships above the Second Death Star was due to the visual effects team getting more efficient with that process of splicing.
On the flip-side, CGI offers many benefits. In terms of cost-effectiveness, let's look at Jurassic Park again. Spielberg made use of both animatronic dinos as well as full-CGI versions. CGI may be expensive, but just imagine how much it would cost to build a properly working, fully articulated animatronic dinosaur - head, eyes, mouth, tongue, arms, legs, and tail. On top of that, the post-production team would have to erase all traces of the supports, cables, and wires used to operate the monster. That would be expensive. The only cheaper route would've been to go with stop-motion or simply use only footage that had been cut around the limitations of the actual animatronic head.
CGI in the Battle of Coruscant is an area where I'd encourage its use. Someone mentioned overkill, but I don't really think that's the case there. Lucas, while a bit excessive, has always been keen for visual detail. He gives us scope and immersion into a monstrous three-dimensional battle above a city-planet, filled with combatants and fire. In terms of physical sets and models, that would be impossible to achieve in any reasonable time frame. Even the best studios run by schedules.
I do have to say, though, that I agree with the OP's overall consensus. CGI - and digital filmmaking in general - has become far too easy to use. It's become a crutch for indulgence, rather than a channel for creating the impossible. That's one reason I'm such a huge fan of Chris Nolan. He uses it where he needs it.
Anyway, I see some benefits to model work. You have a tactile object with literal detailing, imperfections, and character; that translates subconsciously to the viewer that's extremely hard to replicate in a computer. The lighting, while it may be irritatingly difficult to match, is real lighting. However, there are significant drawbacks. The OP posted the Battle of Endor. That entire battle was assembled on the film itself - the physical material - one ship at a time. Each starship was built, filmed, and added on an individual basis. So, while the models may have an edge in reality and may be cost effective, it's also very, very time consuming. The only reason we were able to go from a few X-Wings in a single shot above the First Death Star to dozens of starships above the Second Death Star was due to the visual effects team getting more efficient with that process of splicing.
On the flip-side, CGI offers many benefits. In terms of cost-effectiveness, let's look at Jurassic Park again. Spielberg made use of both animatronic dinos as well as full-CGI versions. CGI may be expensive, but just imagine how much it would cost to build a properly working, fully articulated animatronic dinosaur - head, eyes, mouth, tongue, arms, legs, and tail. On top of that, the post-production team would have to erase all traces of the supports, cables, and wires used to operate the monster. That would be expensive. The only cheaper route would've been to go with stop-motion or simply use only footage that had been cut around the limitations of the actual animatronic head.
CGI in the Battle of Coruscant is an area where I'd encourage its use. Someone mentioned overkill, but I don't really think that's the case there. Lucas, while a bit excessive, has always been keen for visual detail. He gives us scope and immersion into a monstrous three-dimensional battle above a city-planet, filled with combatants and fire. In terms of physical sets and models, that would be impossible to achieve in any reasonable time frame. Even the best studios run by schedules.
I do have to say, though, that I agree with the OP's overall consensus. CGI - and digital filmmaking in general - has become far too easy to use. It's become a crutch for indulgence, rather than a channel for creating the impossible. That's one reason I'm such a huge fan of Chris Nolan. He uses it where he needs it.