Let us assume, for a moment, that you can 'blur' the barriers between online and offline, and that I don't know who I am flying against.

You still need good AI. If there are not enough people online, in the same timezone as me, with free time when I want to fly, then the AI will have to do a good job of pretending to be a human... Thats the purpose of AI!

As Osram pointed out...those who fly online, with friends who do it too, are risking a blinkered approach to say that EVERYONE wants what they want. That is going back to the old 'death of flightsims because' problem. To sell, a sim must appeal to ALL potential buyers, not just a hardcore, online flying simmer. There are not enough of them to justify the work and money of developing. For those who want an online only experience, there are already several options.

Personally, I don't enjoy flying online much. I don't have the time.

Why? Simple. First, find a server...then find one flying the sort of mission/ planes/gametype I want....then go find one that has enough people in it...then go find one without misbehaved children in it...then find one with pilots who are at my level of skill...then

You see the problem

To sell to as many people as possible, a flightsim needs to cater for as many people as possible.

This means:

LOTS of scalable options for difficulty
Good AI that flys planes like they should be flown, with varying levels of skill
Offers stable online play for those who want it
Offers scalable detail to allow average PC owners to play
Lastly..Gives the player the chance to ENJOY THE GAME THE WAY THEY WANT TO ENJOY IT.

Cutting back to the title topic....

I think the main reason for the 'Death' of flightsims is the steady move away from that list, capitalised statement.

The IL2 series got closest, but sims like LOMAC show how wrong you can get it. Only those with super PC's and real flight experience need apply...