That's up to each person to decide for themselves. Nobody can tell them.
That said it sounds like you have a good grasp of the average AAA turnaround in development which is about 5 years.
As for how much further they have to go? That depends on what we are talking about. Star Citizen: Squadron 42 or the Star Citizen: Persistent Universe?
Lets look at just one aspect. FPS.
FPS: Animation and vision stabilization needs to go through a polish (weeks to a month maybe more)
FPS: FPS EVA is missing an entire feature (push/pull) that appears linked to grabby hands. Grabby hands first implementation is coming with 3.0 with the Merchant/Pirate/Bounty Hunter/Mercenary Job. (months)
FPS: Injury/medic job mechanic does not appear to be something slated for 2.6 or even 3.0. It may require a refactoring and polish pass. (months)
FPS: Infiltration mechanic implementation and possible animation refactor and polish (months)
FPS: Character final facial implementation including customization (months).
FPS: New netcode. Current netcode is not viable for a released game. StarNetwork 1.0 is slated for 3.0. It will obviously require testing and polish. (months).
This could be cumulative in nature. Beginning to see a pattern?
Now some of these might be implemented at the same time but lets assume worst case scenario and it's in waves which requires time for a full polish pass before it even hits PTU. CIG may hold entire features back if they require too much animation rework and do them all in one go to save time but that may hold entire features back from player testing and nobody would see progress for many months.
Depending on which of any of the FPS features above is required for SQ42 that will push SQ42 release back.
Here are things they will absolutely need for SQ42 single player. The rest we aren't 100% sure:
1: Final Animation and Vision Stabilization
2: Infiltration
3: Final character facial implementation
I left Injury/medic off since that's a question mark at this point. EVA push/pull will be needed for the PU but they could do SQ42 without it and it's not clear it's absolutely necessary. Same story with Grabby Hands. Netcode is not required for SQ42. It will unlock multiplayer Co-Op major missions from the SQ42 story you can play with friends and for that you need netcode but you can release SQ42 without it.
Of the above 3 they are closest with #1 for completion. Ivo Herzeg who came up with much of this tech and is implementing it had an interview with Brian Chambers (Director at the CIG German office) and based on that I would say they are very close.
You can watch that here.
FPS Infiltration we know nothing about other than it's a requirement for SQ42 to launch.
Character facial implementation they talked a lot about this week and was a feature of this weeks 10FTC Special.
Here is an earlier video on it from 2 weeks ago as background information so you can see where they are.
Here is the 10FTC Special that deep dives into characters.
Based on the above there is no reasonable way they can make a SQ42 release for 2016. Even if they completed all features midnight tonight and the entire community did nothing but testing between now and mid December when CIG goes on vacation it's not enough time to give the game a "Naughty Dog" polish.
Knowing this and knowing that SQ42 will no doubt launch before the Persistent Universe does there is no Star Citizen in 2016.
Star Citizen has been in what most AAA developers would consider development for just under 4 years. Also most AAA developers have existing employees, culture and most people know each other well enough to be approachable to solve problems as a group instead of working against one another which happens in the outset. Basically people know what to expect.
It's easy to compare budgets with 20/20 hindsight without taking everything else into account. It's professionally called growing pains. It's not so easy to compare CIG to other studios outside of the budget. The closest comparison I can think of is CDPR since they grew from around 50-60 employees to well over 200+ for the Witcher 3 and ended up having to sell out a bit to afford that huge influx. But even they had a core development group that worked like clockwork.
CIG hasn't had any of the sort until the German office was formed in very late 2014 with just a dozen plus people in practically a closet at first but these people worked together in many cases for 10+ years. What a difference.
It's up to each person to decide if you want to give them a little break or not while they caught up as a company.
I don't know when SQ42 will launch. With only two major features missing yet one seemingly well in hand but the other completely unknown it's impossible to judge. 2017 seems reasonable but when I wouldn't trust any specific dates even the ones CIG will release at CitizenCon. Remember what I showed above. It's not that simple to get from a tech to release. It takes money and time.
Let me repeat that. Do not put any hope into release dates given by CIG at CitizenCon or any hints at release dates.
Final reminder. I brought up FPS only. I did not bring up other systems that could further delay SQ42.
And this is why it's hard to reply to generic questions when you know a lot about something.