You are technically right I'm prematurely assuming there will be bottlenecks. But as almost every game seems to need optimization these days, that brings up the question, why do they need optimization? From looking at some games, (i'm staring at you bethesaida), these games seem to lack a consistent standard of quality throughout development and thus require the optimization. Though, in Bethesaida's case, you'd think they would know by NOW how to do a proper open world game, as Fallout 4 is their 5th mega-large open world game (haven't played any of their's before morrowind so only counting Morrowind, Oblivion, Fallout 3, Skyrim, and Fallout 4). They know their engine and their file formats back to front... why do their code have so many bugs? ... If you were an employer you'd expect that over time and over several projects, they would become more and more experienced, such that their quality goes up. With bethesaida, it doesn't seem like they learn anything as some of their bugs from Morrowind era, are still there in Fallout 4. I dunno, I expect bottlenecks, because many triple AAA games seem to run into bottlenecks and as a result require huge optimizations.
However, from what I do understand about optimization, its often things that developers have little control over directly, such as reaching your draw call limit. I heard that before Assassin's Creed Unity was patched it was hitting 25000 draw calls per frame, when the maximum they could do per frame to maintain a stable framerate was 10000. How does one optimize that?
Well, from looking at the documentation, I could create a relatively basic UI, and code specific commands to keybindings, instead of buttons on the interface. I had intended to create and test several basic commands, such as Move, Stop, Attack, and Stand Ground. I wanted to test out the hitboxes to ensure that weapons fire was indeed hitting the target and causing some kind of effect (ie loss of HP as a very basic example). I wanted to ensure that when a unit is destroyed, it leaves behind a husk; similar to those husks that are created in Supreme Commander when a unit is destroyed. Reflexive AI would also be something I'd want to test out; it would be important to test it out with a command like Stand Ground, to ensure that the unit will stay where it is but attack at enemies within range, but units without a Stand Ground command would need to move within range and attack.
I've spent some time looking at how keybindings work in UE4, and how to setup a way to rebind those keys. I've taken some time to also look at UMG, the tool used to create interfaces for UE4. I've messed around a bit with camera controls, to get the camera to the right angle and height for a RTS. You can create objects made from prefabs, and attach code to them and thus use them as temporary objects for units until the actual models are ready. From what I've learned from it most of the stuff is straight forward, but I still need to know how to setup the gamestate, UI/Camera and player controller for a RTS. I'd want to reread these sections again though in case they have changed much since 4.7 when I read them last (they've released 4 patches since then, with a possible 5th coming out soon) and as many patches have come out since then, its likely they have. A good place to start would be to check out, as suggested by someone here, the tower defense game demo that UE4 released for the engine and see how it does things to get an idea.
