The COFFIN system is called so (c'mon, there is no way this acronym wasn't intentional!) because the pilot lies on his/her back, closed in from all sides. In modern planes, the pilot is usually seated to give them a better view and more intuitive controls but they are also more prone to blacking out from excessive G-force, because human body can take much less Gs longitudinally than laterally. Now, in an advanced airspace display like COFFIN, there is no need to have the pilot seated to give him/her 360 degree view, while laying them on their back would allow them to pull much sharper brakes and turns! The matter of ejecting from a COFFIN is still open, however... unless Fridge Horror applies here, as well. — Koveras
Fridge Horror definitely applies here. All of the benefits to the COFFIN system above? They come at the cost of having no ejection system - if you get shot down, you'd better hope it's a smooth crash-landing.
It gets worse. To use the COFFIN, the pilot's brain must be connected with the Electrosphere. That leaves him/her vulnerable to hacking. Something that is used with extreme prejudice in most of the late-game missions.
In real life, it's currently believed that fighters are currently at the limit of what conventional manned aircraft can achieve, not due to aerodynamic or engine limits, but the pilot themself. One way around this is the above, which was actually briefly experimented with in "prone pilot" aircraft, before being discarded due to hampered visibility and other hassles. The other way of getting around the pilot limit is to just get rid of the pilot, and use an AI instead. The main drawback is, of course, the difficulty creating a robust AI that can truly match the skill of even a novice pilot, but the advantage is that the only G limits that matter are those that exceed what the airframe itself can handle, and perhaps the greatest advantage is that when you lose a plane, its a hundred million dollars down the drain, while a good pilot takes years of training, has to be trained from scratch if they die, and they will eventually grow old and lose their quick reflexes. An AI would have no such issues, since it could easily be, "saved" at base, then loaded on another aircraft if the first one is lost
The Omega Ending reveals that Nemo is an Artificial Intelligence, and Skies Unknown all but spells it out that its a descendant of the ZOE. Considering how the ZOE evolved through the franchise, you just wonder how powerful it got. A good idea would be from how ZOE was able to quickly learn from Scarface, then Hugin and Munin nearly causing a Robot War while learning even more quickly from Trigger. Then you finally have Nemo, who can learn maneuvers in an instant (you start the game right off the bat with your squadmates asking you to stop imitating their flight style) and take on entire air forces of superplanes. Just imagine being the victim.
What makes this even worse is how Nemo operates. Gathering intel on the situation either by searching in The Alternet for news, files and whatnot, or by hacking cameras to observe firsthand what's happening. In a fight, when it notices that the enemy has the upper hand, either by having a better plane or being a better pilot, it decides that dogfighting alone isn't enough and starts resorting to underhanded tricks. Rena was the worst victim of this. In one of the late-game missions, Nemo hacked its way into her brain and tore up her mind, only deciding to shoot her down after she became nothing but a screaming and crying mess. That shows that Nemo can learn more about war tactics than flight maneuvers.
In other Ace Combat games the final missions usually have an awesome music score fitting for a climatic battle. However some final missions in Electrosphere have a really unsettling track such as Zero Sum and Morceaux, which wouldn't stand out in a horror game. Those tracks appear when Nemo's true nature as an AI is greatly hinted at, making the Player Character one serious Nightmare Fuel.