EPTI think 1v1 Competitive Brood War should be regarded as a Siege right from the start.
It may end quickly with a raid, a frontal attack, or many sneak attacks.
However I think the game will only truly reach a strategic pinnacle as players figure out how to either block these attacks or find attack opportunities of their own to prevent later attacks.
Consequently, it is my suspicion that increasingly the game should be viewed as a war-siege rather than as a skirmish or battle simulator.
Some concepts I think are brought into more prominence by this framing:
1) Logistical superiority - early on resources are tight and great care must be taken to prevent the opponent from getting an overwhelming lead in logistics and supply. Later on, the attacking force tries to overwhelm the defending force with superior logistics and can afford to be less efficient when it comes to breaking the weaker enemy.
2) Specific operations - identify the different operations needed in the Siege as distinct elements of optimization. How do you fortify your castle? How do you march? How do make camp? How do you conduct a siege? How do you conduct a raid? How do you engage an army in open field? Each of these deserves some thought beforehand so that in the fight you already know your best option(s). Ideally every one of these optimizes both for tactical resilience but also efficiency of resources and APM.
3) Time - Once you get past the chaotic opening stages of a match, you'll have a lot more time. Fights in Brood War often take minutes. You do need to be APM efficient, but you can make plays that take longer and longer to pay off, such as having some SCVs in a control group and shift-right-clicking through a mech position after a fight. That repair will take a long time, but you'll have a long time, unlike in the early game.
Here are two replays which I believe show this understanding coming into clearer focus
Larva loses to a siege with constant pressure
Light conducts the siege from a logistics-first perspective, showing a more dominant result
I think one consequence of this framing is that you start to move away from the battle simulator mentality.
You don't have to "Go Mech" or "Go Bio" so you can get the biggest army and crush them in the field.
That's an option, sure, and maybe sometimes even the best option.
But what you really have is a bunch of strategic objectives on the path to some sort of victory in the siege (which might be a raid 2 minutes into the game or it might be taking half the map).
At each point, you want to select the units that contribute to your strategic goals in the siege process, whether or not they "go together" in some abstract battle simulator way.
Yes, resource requirements are real, and yes you don't wanna just get crushed early on, so you need to be efficient and work your way up to have the freedom to get the tools you need. But you definitely want to get all the best tools for the tasks at hand.
An important clarification: how to make camp, how to siege, etc etc are plans. Layout matters but they are not just sceenshots of some layout, but an order of things to do and a tempo to do them. Maybe once you're at a certain point in the siege you send in harassment to overwhelm him. Maybe as you defend a camp you gradually add turrets.
Moreover, these plans have to be a little flexible to account for the enemy's choices.
You'll likely have to develop an intuitive sense about whether to commit a raiding party to the enemy's gates.
You'll have to have a sense of when you need a little more or less robust fortification than normal.
And so on.
So there is a whole maze of judgement calls to make here, and tools to punish the opponent for judging wrongly.
