No worries. [MENTION=5834]Celtavian[/MENTION] obviously doesn't need anyone to white knight for him (!), but some of the comments being made seem to me to be way off base.
I'm a great believer in actual play posts, rather than abstract speculation, generalisation or theory-crafting, and Celtavian has provided a pretty clear actual play post that explains how his group defeated this marilith + friends encounter pretty straightforwardly (with an admitted misreading of the Banishment spell - we don't know how much difference it would have made had the Concentration requirement been imposed).
I think we can infer from the extensive use of other concentration spells that it would have made a significant difference. I'm not saying they would have died or anything, but I think it would have been noticeable.
I think you're right about different viewpoints. Certainly, from the optimising point of view the suggestion to sunder the PC archer's bow is bad in so many ways, as Celtavian, Hemlock and others have pointed out.
I don't think my group is at Celtavian's degree of optimisation, but I've had other posters express surprise at the number of encounters (including but not limited to combat encounters) that my 4e party can get through without an extended rest. So I think, either in individual build or in synergistic play, my players are probably somewhere near the upper end of the spectrum.
We've been lucky, I think, that no super-obvious "win" options have emerged in our game (nothing analogous to the ranged attack dominance in 5e, or the darkvision shenanigans), and 4e provides good support for complex monsters, tricky tactics and interesting environments. (I know Celtavian hasn't played much 4e, but I think his group could enjoy it, by the sounds of things.)
Your prescription for 5e seems sensible. And Celtavian has said himself that he wants to dial back Sharpshooter, which seems to be one of the biggest issues. (Maybe also make Close Quarters and Archery either/or, rather than stacking?)
My experience with 4e optimization (using lots of splat books, themes, the whole nine yards) was actually a lot of fun. A ridiculously tanky fighter (he gained temp hp every round from his core ability, I forget the details) who had buffed his basic attack to ludicrous levels, coupled with a warlord and a bard to grant him a lot of extra attacks. Toss in a strength-heavy monk for crowd control and as a backup target of melee basic boosting, and they made a hash of almost every encounter despite lacking much ranged firepower. 4e had its moments, even though the fundamental meta game ran counter to everything I like most in an RPG.
Anyway, yeah, even with the un-optimized 5e parties I've been managing lately, I've noticed ranged dominance. It seems pretty wonky to me, in some respects. For example, the ranged fighting style is the only one to give an accuracy boost? Really? So shooting a guy with a bow is easier than whacking a guy right next to you?
Also, I'm not too familiar with UA, but from the bonus breakdown it sounds like CQC grants a +1 to hit? Seems like a terrible idea on the part of the designers if so.