Sure people over-complicate, but the present example illustrates an opposite tendency to oversimplify. Bad anchors, questionable moves and pro ahead---how the hell is the party gonna back off? And what if, for example, they are way up on a wall in a remote setting?
I think this situation is one of the most nuanced---and hence complicated---a party can face.
I guess the other thing alluded to here ( thanks again to the author) was that all belays should be created for easy escape and instant conversion into lowering / rescue rappeling and ideally even raising. I certainly can get a bit lazy especially with double ropes but when the sh#t hits the fan - which is why we rope up in the first place - you need a single bomber point that you can instantly escape from, especially if your partner is entering the "golden hour" zone of getting to surgery ASAP.
Minimize clusterf*#k at the station by building it out of the minimum amount of biners and slings. The simpler the anchor piece interconnects are the better. Make 2 tie in points easily unclipable with large biners like Petzl Attaches. Orient to the direction(s) of predicted pull, including up. The simplest belays are made with just the rope, and 3 good pieces attached by only 3 or 4 biners. This includes you Big Mike ...
Sliding X always works because you load tested your primary anchor points first to be certain it's not going to blow.
Carry a small knife in your pocket and spare rap slings/cord in the pack. Carry two short prussiks on your harness for managing a weighted rope and for backing up your rappels.
Pure RGold to be found here, love the advice and discussion.
I keep the "Climbing Anchors" book by Largo and Bob Gaines by my bed for frequent reading. I take forever to build an anchor - out of paranoia - and none have ever failed, but then none have ever had to hold a fall..... Yes, they have held for TR and rappel.