What factors make armor a good idea?
“Urban” or “Low Intensity” activities: hallways and doorways, mainly human-sized opponents, squads rather than battalions. The bigger the armor, the more cumbersome, however. You can’t realistically climb stairs if your feet are too big.
No A.I.'s that can adequately do the job, and no tele-operation is feasible (the drone is the killer app). Another killer app: genetic engineering of “creatures” (not necessarily very human) that can get the job done.
Adequate power and mechanical tech.
Difficult environment (atmospheric composition, pressure, maybe vacuum, NBC situations). NBC = Nuclear, Biological, Chemical weapons.
In the excellent webcomic Schlock Mercenary, the tech available make it feasible to build pretty tough armor into a regular uniform. No weapons or strength augmentation, but a good idea. The weapons are guns, rifles, and tanks. This makes sense, since guns may need a lot more power, can give off a lot of heat or exhaust or have a mean recoil, and allow for more variety with a minimum of fuss. Do you really need superior physical strength at the cost of agility? Do you need big bulky armor with lots of guns stuck on it, making you a bigger target for others?
In Peter F. Hamilton’s “The Naked God” books, mercenaries and soldiers received surgery and implants and augmentations that gave them their edge: chameleon skin, boosted strength, “wraparound” or extra eyes, armor, connection to communication gear and computers, extra arms, all-around puncture- and burn-proofing, ability to survive many G’s of acceleration, etc.
A lot of thought went into some of the Iron Man tech, some good ideas or maybe inspired post facto invention. The latest armors are a sandwich of different layers with different functions: various armors on the outside, a power source layer, a servo-motor layer, a layer to sense the wearer’s movements, environmental conditioning, the computing elements, etc. Everything at a very small scale, making the armor less bulky. The War Machine armor (the black and silver one) carried a lot of firepower (rockets, flamethrower, chain gun) that achieved a big force multiplier effect, but also gave a lot of variable response flexibility (e.g. non-lethal force, short range and long range, armor piercing, etc.). The multi-millionaire superhero schtick does away with a lot of the questions we’ve brought up, like cost and motivation.
Starship Troopers is a good read and the armor had some clever functions, but it’s based on badly outdated and unimaginative concepts of warfare and strategy. The Starship Troopers cartoon had a good idea: a squad of marines was augmented with one or two armor-wearing soldiers (although the armors were close to being tanks): force-multiplier effect, gave the unit more options and a bigger punch.
The Sand Wars books mention one interesting use for armored operatives: body guards. On the other hand, can you imagine suicide bombers wearing some of these puppies? Hopefully, that will never be anything but wild speculation. Think of RoboCop, on the other hand. Nearly armor, and there aren’t big logic or plot holes. If the makers offer the tech to amputees or people with birth defects, they’ll get a whoile hell of a lot of volunteers.
Armor won’t ever be a good idea if the guns keep getting deadlier and their range keeps increasing. Picture armored soldiers against automated drones that pick up heat or electronic signatures. Or just motion sensors. It’s way easier and cheaper to make smart mines and big guns (or big tasers, to short out circuitry).