The thing is, I think its fairly obvious they had some major re-write or had to go back and re-do something fairly important that it added an extra two years.
I think this is an accurate assessment. I talked to another (much smaller) game developer who also switched from Cryengine to Lumberyard, what he said was, "It's basically compatible, but there are very time consuming issues" without elaborating.
Both AWS and CIG had forks of Cryengine to start with, forks that were from different versions of the software. This kind of thing, what amounts to changes of software version mid-project can be fatal to a project. I've seen it many times.
Also, I think that CIG has a lot more developed than they have stable, but in some ways they put the cart before the horse. If the PU is dependent on SQ42, then that should be the development priority - but they don't seem to get paid very much for that.
AND I think most of the people who thing SC in general is taking too long, are mostly just impatient. They want to play the game NOW, not in 1, 2, 4 years. When the project was originally announced it was incredibly ambitious and forward thinking. This can be a bit of a problem, deciding what to put in the game when the technology isn't capable when you try to develop it. Then changes occur, and the technology is capable but not in the way you originally planned. I suspect, this was one of the primary motives for their relationship with Intel.
Yeah I still think this is really a 10 year project. When i look at other games from around the same date as the kickstarter, I don't see particularly more development. Elite Dangerous maybe, but they already had a platform - and they concentrated on simplistic mechanics, but produced a playable game. This is another area where CIG may have made a mistake: It's OK to offer game updates, after you release the basic software. But it is what it is at this point.
