The core idea I teach my students is that choices in a game should be contingent and situational.
That is, there should be situations in which you would choose one option and situations in which you would choose another. This leverages the player's skill, judgement, and intuition in discerning which of these situations they're in.
You give an example of this with head vs body shots.
"My hit chances are low, I'd better aim conservatively to minimize the risk of missing" --> aim for body
"My hit chances are high, I'm not at much risk of missing and can prioritize damage" --> aim for head
Here, the exact threshold between "low" and "high" hit chance, or how much risk of missing the player is willing to take, are subjective and situational. Different players will come to different conclusions in different situations, depending on their aggression or risk tolerance or details of the scenario.
Maybe the enemy is on its last legs and so damage isn't so important as just landing a hit to end the fight. Maybe you're on your last legs and if you don't get a headshot now it's over, so it's an all-or-nothing gamble. Maybe you're concerned about running out of ammo, maybe you're more concerned about draining your healing item supply. Each of these different elements add layers to the decision, and opportunities for players to choose to prioritize one or another.
Together, these considerations keep the player's mind and subjectivity engaged in the decision, so it doesn't just collapse to a calculation.
Another important aspect of contingency is beliefs / predictions about your opponent's strategy, or unseen information like what's around the next corner. Even if your opponent is an AI, knowing that they have, say, a 20% chance to do a berserk charge when reduced to 10% health will impact players' decisions about when to prioritize damage or reliability or other effects like disarming/stunning/immobilizing/knocking back.
You can take pretty much every system in your game and brainstorm actions where it would be a positive or negative factor. Got a loot system? Maybe some actions can boost your loot drops or increase the chance of rare items, so I'll play differently when farming/grinding for a specific drop. Got a reputation/morality system? Maybe some combat choices are more honourable than others, so I'll play differently depending on the character I'm roleplaying or the factions I'm trying to befriend.
Each time you weigh adding a new combat option, the question to ask yourself is "in what situations should a player choose this over the options I already have?" and "in what situations should a player choose another option over this one?"
If you have strong answers to both questions, that's a sign the choice will create interesting decisions for your player. Particularly if sorting between those situations demands considering lots of factors (like my health, its health, engagement distance, amount of ammo, healing item supply, distance from a safe location, chance of backup, etc...) that can test their observation, quick thinking, tactical judgement, resource management, priorities, and intuition.
If there's only a few situations where it pays to choose one thing or its alternatives, or those situations are obvious (lacking the subjective assessment of high/low and risk of the example above), then it's likely to collapse to a calculation instead of challenging your players.
There's a good Extra Credits video on this topic too. 😊