Preaching to the Choir
As John points out, we have more advanced materials today- plastics that can withstand greater impacts and temperature extremes, more stable rubber compounds, advanced plating techniques, etc. However, the reason our classics run circles around the new machines is obvious- the bottom line is manufacturing cost, not advancement. New machines are built to give the consumer exactly what they want- a cheap machine that looks expensive.
The number of '06 series Maytags humming along in midwest basements, decade after decade, is a testament to the "robust design" philosophy that was part of the industry. We all now there is simply no way a manufacturer could produce a machine like that today. Switch assemblies with individual springs and terminals, screwed into place rather than a slot and rivet method. Pushbuttons with distinctive part numbers (inventory and QC), a porcelain tub (multiple processes to one part, and costly), a motor with line-voltage rated windings of adequate horsepower. Steel case stampings of adequate gauge that they don't require reinforcing ribs. And so on.
Plastic parts don't require painting, they can come right from the supplier, be snapped into place on the front of the machine, and the sticky-backed, multi-button overlay with graphic pressed into place. No tools required. A single PCB with membrane switches and processor mounts from the back. Maybe there's a couple screws in there, but if manufacturer A finds a way to do it with one less, competitor B will find a way to beat them.
Of course we're dealing with inductive loads like water valves and motor windings. Interfacing with low-voltage digital electronics means we need "protective" components in the interface. These can often cost more than the processor and software (yes, there is firmware in there) running the machine. Just which parts do you think get marginalized first as a cost-cutting measure? Put all this in a damp environment, add vibration, and put on your EE hat. The list of potential component failures that can turn that machine into a paperweight gets pretty long. Contrast this to the old model - timer contacts and a motor with capacitor start. That's it. A failure is bound to be mechanical because the electrical system is up to the task.
We have GE to thank for value engineering, but the impetus was manufacturing during WWII. Advancements in technology allowed design to evolve to the point where systems and components did not require "over building". If a stamped bracket will survive 5 years of service before metal fatigue causes failure, why are we paying double to cast this part? The problem is, instead of using this approach to optimize design, manufacturers used this approach to build-in obsolesence. We're now in a race to the bottom.