There is nothing wrong with fly-by-wire and full authority controls as long as it is done correctly, it can even be more reliable than traditional mechanisms.
The way these things are made reliable is by redundancy and formal testing. The Boeing 737 Max MCAS had neither: it relied on a single angle of attack sensor and the software wasn't certified up to the appropriate level. The big mistake was that they gave the system more authority than what it was originally designed for, without an appropriate requalification.
Exactly. MCAS had only one goal: making the Boeing Max feel like the old Boeings by fiddling with the trim so that expensive retraining could be avoided. A fly by wire solution would have resulted in more design changes in the cockpit thus requiring a new type rating for lots of pilots.
IMHO it is entirely fixable technically but of course the scrutiny of the FAA on this and the resulting steady flow of management fuckups, more potential issues being unveiled, and the apparent failure of the certification process is pretty much guaranteeing this to take quite long. Last year I was still optimistic it would fly again soon but given the recent trickle of more issues, I think this is going to be a pro-longed grounding.
Airbus had it's fair share of incidents involving fly by wire in the eighties and nineties. However, more recent incidents seem to be not related to that anymore (terrorist attacks, bird strikes, bad weather/decision making, etc.). So, fly by wire is entirely safe if done properly these days and also not optional for flying most modern military planes. Airbus basically laid the foundation for their current success in the market right then.
IMHO pilots are increasingly becoming safety pilots (i.e. emergencies and unusual situations are the only times they fly manually) and we're not that long away from fully autonomous planes. Right now, aside from takeoff and landing, most passenger planes are controlled via the auto pilot only. The pilot doesn't touch the yoke and instead fiddles with buttons to direct the auto pilot. Cat III landings can be fully automated technically and certified pilots are required to fly a certain amount of fully automated touchdowns regularly. The military routinely flies remotely controlled drones and fully autonomous drones already.
So, a design like this could make sense. 20% reduction in drag is quite a lot of gain and fly by wire has been used for decades to make all sorts of otherwise unstable configurations flyable.
The way these things are made reliable is by redundancy and formal testing. The Boeing 737 Max MCAS had neither: it relied on a single angle of attack sensor and the software wasn't certified up to the appropriate level. The big mistake was that they gave the system more authority than what it was originally designed for, without an appropriate requalification.