A point of view that has come up a few times recently is what's been referred to as "the path of least resistance".
I believe the path of least resistance is only problematic when players are competing among themselves, at which point using any means at all to improve the performance becomes "required". Otherwise, the easy path is often used to bypass things the player doesn't actually like, which is good - which is why many games even include some kind of limited use super-weapon that allows the player to bypass part of the game if he desires.
Of course, MMOs have a tendency to have the players competing among themselves all the time...
Regarding the cockpit/HUD debate, I think that there are muddying factors that make the case less clear-cut. Some people might find the cockpit ruins their immersion, or struggle with it in terms of eyesight/motion sickness. However, people who want the cockpit might find themselves abandoning it if it puts them at a significant disadvantage. The latter example is about the path of least resistance, but the former two aren't.
IMHO the best option in this specific case is to try to build into the cockpit view enough resources to even out the field. If the third person view has more spatial awareness, build into the cockpit view features to make it easier to detect anything important around the ship, enough to mostly, or completely, nullify the third person view advantage.
Not only this should solve the problem, it's actually very realistic, given that whoever designed the ships in-lore would have attempted to provide this kind of advantage anyway (at least if he wants to sell his ships). After all, if we will have holographic displays for other kind of info, anything that the devs can think of to increase the spatial awareness inside the cockpit could effectively be created with the in-universe technology.
A lot of work has been done on managing real desire paths - figuring out how to detect and support them where possible, and how to redirect desire where necessary. Here are some solutions they use that might work in gaming:
- Measure what people actually do, then just put the path there. A/B testing is all well and good, but there's no substitute for old-fashioned observation
- If you can't put the path there, put something horrible there instead. A path with a rope keeping people in just makes it worse, but a path with a thicket of weeds either side makes people feel safe
- If you can't put something horrible there, educate people about the problem. Knowing that a single large group can damage soft wilderness ground beyond repair is enough to make most people do the right thing, so maybe knowing the cockpit is there to increase immersion will be enough to make us use it
I don't think the education example in the third solution is actually appropriate.
First, because saying that a player must use a certain view mode, which theoretically only affects himself, due to immersion is silly. As long as I'm not harming others, I will play in whichever way I find more fun, which can potentially include doing things that break immersion for myself. (I would avoid breaking immersion for other players, though, because I do see that as potentially harming them, or at least their experience.)
Second, because if it's something where a fair number of players engaging in the frowned upon behavior can make the game worse for everyone else, I don't think education alone is nearly enough. I think the education approach only proper for when you are still quite fine with a reasonable number of players doing the "forbidden" thing.