I have no idea. What are your benchmarks for those? What are your performance needs?
“complex boss fight patterns” - absolutely.
“multiple pool of weapons” - Lots of weapons? How many is lots? Borderlands 1 had something like 17 million different possible weapons. Borderlands 2 and the Pre-Sequel have even more. A quick count from this page shows nearly 80 in Call of Duty: Ghosts. Unless your weapons are going to be generated completely from scratch, including graphics, you’re going to make blueprints for them anyway, even if all your logic is in C++.
“that interact differently with the enoviorment” - Depends on how deep of interactions you plan.
“RPG text dialogue…system” - Yeah, this is feasible in Blueprints. I would write it in C++ for organization’s sake, but you don’t need blazing fast performance for a dialogue tree.
“save system” - Again would probably do this in C++ for cleanliness, but if you’re really dead set on not learning/using C++, sure.
Why do this in blueprints? Why not just have textured models?
From a quick glance at Google results, that game might surprise you with how much they have to do to achieve that art style.
If you really know what you’re doing, C++ will be a little faster for the things you’ve mentioned, but not so much that making them in Blueprints would be “wrong”.
If you don’t know what you’re doing, neither C++ nor Blueprints will be optimized enough for you.
The question is not “Are blueprints optimized enough?” The question is “Can I pull off a game like this with my current level of skill?”.
When I say this I don’t mean it in any offensive way, but it sounds like the answer to that question is “no”. And that’s okay, because your current level of skill can change quickly if you take small steps. Follow the tutorials in the learn tab. Make your own game, as absolutely tiny as you can come up with, and you’ll be surprised at how much work it is to polish it enough to be comfortable sharing it publically.
Why do I assume “no”? Because you’re asking vague questions. If you were asking concrete questions they’d be a lot easier to answer and you’d be a lot closer to being able to realize what you’re working on. Learn enough to ask concrete questions. Then learn more. And soon you’ll be able to tell whether you can accomplish what you want. (or at least be able to create just enough to test the viability of an idea.)