Future Flexible Armour
The most likely solution to this is ‘momentary inflexibility’. Whether this is achieved by magnetising the suit in anticipation of the impact, or through some clever use of dilatant fluids, the principle is the same. The suit is flexible except when being struck, at which point rigidity lasts long enough to resist the impact.
One way to stay mobile while also being indestructible is to rely on some sort of energy field that repels incoming projectiles. An advantage unique to this sort of defence is that it makes no differentiation between the ‘joint’ areas and the ‘rigid’ areas of the human body and this makes it a huge boon to anyone in critical need of dexterity (brain surgeons take note). Unfortunately, whether this is a plasma shield, or a magnetic screen or whatever future science you choose, it does still have some limitations.
Firstly you need to think about how your shield user will interact with the world. How do they breathe? What if they need to pick up something outside of the shield, or shake hands with someone? How does your shield differentiate between a dangerous projectile and the air, or the floor underfoot?
Short of rewriting physics there’s not really a ‘solution’ for this problem but its also very similar to the weight problem faced by conventional body armour. By reducing shield coverage, you can apply more of your power source against the threat. Further more it may not be necessary to make a shield that can stop every kind of weapon. Just because you can acquire a shield powered by antimatter, doesn’t mean the person attacking you can acquire a weapon that will go through it.
Most world builders will be familiar with the idea of powered armour – take a suit of armour too heavy for a normal human, and then use servos and motors to drive the limbs so the user isn’t fatigued. In some universes the powered armour also provides tremendous strength to the user, so they can throw enemies through walls and such.
The biggest obstacle to powered armour is extremely close to the heart of most military folk, but rarely considered by most gamers – serviceability. Serviceability is how easy a machine is to maintain and how likely it is to breakdown, even if well maintained. A basic human rifleman has an almost perfect serviceability ratio – with sufficient need they will find a way to fight even if tired, hungry or injured, even if their normal equipment is damaged or missing, they’ll find a way to engage the enemy.
Powered armour complicates this a lot. If a unit is marching overland and one of the suits breaks down what does the unit do? Abandon the suit and have one member fight unarmoured? Wait in place until a repair crew can arrive? Leave the individual to protect their suit while they press on a man down?
Obviously in your future it might be predicated that things break down less often but it might be unrealistic to say ‘never’. Consider that a basic rifleman relies on one moving part (the trigger mechanism in their gun) while a powered armour suit will have some amount of machinery in every joint (at least fifteen) and an electrical system to drive it all. All else being equal your powered armour suit is fifteen times more likely to break down than your rifle.
It’s also probable that this armour takes some time, and perhaps equipment, to get into. This means that if a base is attacked unexpectedly (and enemies so rarely RSVP), the defenders are likely to have to fight without their armour.
All of this is not to say that powered armour wouldn’t exist. Attack helicopters also take quite a bit of time to get ready and can have poor serviceability rates, but most armies won’t leave home without them. The key detail here is that an army is unlikely to hand out powered armour to every single soldier. Most likely specialised ‘heavy infantry units’ would use this gear and such units would likely be responsible for attacking the enemy, rather than holding ground.
Now its entirely likely that most designers would choose to provide a little of each, but there would absolutely be variants for every niche. There would be light exosuits that provide tremendous physical strength and agility but no additional protection and equally there would be dreadnaught suits that are no faster or stronger than an unaided soldier, but have higher levels of protection.
It's also very likely that powered armour suits will follow the present-day tradition of focusing armour coverage around the torso. The issue of heavy weight away from the core wouldn’t just fatigue humans, it would stress mechanisms and deplete batteries. For the same tech level and investment you could have a lot more torso armour than you could have protection on the extremities.
war... war never changes
The common thread here is that many of the underlying challenges armour manufacturers face will be as true in the fusion age as they were in the bronze age. Foremost among these is that, for a given weight and technology level, a weapon will always defeat an armour because the armour must express defence in many directions while a weapon need only attack a single point. Short of a system that can predict incoming attacks with precognitive accuracy it is unlikely this will go away.
As a result of this limitation armour manufacturers will likely continue to focus their protection on the most vital regions of the human body. Their focus will remain ‘saving lives as a last resort’ and not ‘creating a bullet proof superman’.