Not saying that every attack shouldn't have gone through him given what you say about BWP/Royer/Veron. Just that the high assist numbers - both actual/hockey and primary expected ones - are heavily affected by the fact that he was the primary and often the lone creator. I think it's fair to mention that context when you or others say he's a league leader in those categories.
Sacha Kljestan by year (Statistics all courtesy of American Soccer Analysis.)
xA 8.89 KP 80 Touch % 11.7%
xA: 9.97 KP 105 Touch % 10.9%
xA: 11.33 KP 111 Touch % 12.8%
I'm not really seeing what you're talking about here. He had a bigger share of the team's touches in 2017, and his xA and key passes went up at a rate pretty consistent to that. No, he wasn't as good in 2017 as he was in 2016, but he was still really damn good. To compare, Nicolas Lodeiro had fewer key passes and expected assists than Sacha despite playing more minutes and having a bigger share of his team's touches.
He doesn't offer much when pressing anymore or with his shooting (he has a good shot but for whatever reason refuses to pull the trigger). We've tried and failed with a team where a younger/physically better Sacha is the best player.
You could argue that adding better wing players and better defense, or better coaching or even just a bit more luck could have put us over the top in any of the 15/16/17 seasons (and I would agree!), but looking at 2018 and beyond, I don't think that Sacha being our best player will take us where we want to. Time to trade him for a decent haul, especially given his value will only go down from where it is now.
Thierry Henry wasn't as good at many things, such as pressing and finishing, when he was 37 compared to 33. So he adapted. Trading Sacha in itself doesn't make the team better. It makes the team significantly worse. I'm confused as to why that's your solution instead of to add more good players.