Obviously you know better than I, but I was under the impression that the T7 and later had a more software-driven architecture.
As I noted, the Boeing autothrottle algorithms haven't changed much since the 767/747-400 days - running closed loop on EPR/N1 Command. The biggest difference between the 777 and the 747/767 autothrottle is that the 777 can move the two throttles independently (via separate servos - 747/767 are single servo autothrottles, meaning all throttles have to do the same thing).
The 747/767/777/787 all make use of a limited 'trimmer' - basically the FMC/TMC sends a signal to the engines telling each engine to move its N1/EPR up or down a slight amount to align EPR/N1 across wing. It's basically there on the 747-400/767 to account for small amounts of throttle stagger. On the 777 and 787 the trimmer is pretty much redundant since the throttles can be adjusted independently by the autothrottle. But the key word is "limited" - the trimmer authority is limited to ~2.5% thrust at low altitude (~5% at cruise). On Boeing, the throttle position determines thrust - a fundamental difference between Boeing and Airbus - one that I'd be very surprised if it ever changed because Boeing considers the moving throttle position feedback to be safer.
As for back-driving the control yokes, my Boeing job is control systems. Artificial feel isn't hard - it's basically a few relatively simple algorithms and a bunch of table lookups (back driven steering wheels and control yokes have been common features on the more sophisticated videogames for years). The hard part is getting the data tables right
But back to the original subject, I haven't had much involvement with the 737NG, but I'm reasonably sure it has a dual servo autothrottle - something that I'm sure will be carried over to the MAX.