In the land of “Modern digital electronics aren’t the only way to solve problems, and clever mechanical engineering solves an awful lot…” - I present this toaster.
I was just reading that article from another source’s linkage. Very neat design and quite clever. Today they probably wouldn’t even bother thinking about it until they’d modelled it using any variety of crazy software, I’m sure. Back then they just mathed it out using well understood rules of thumb and experienced intuition, and then made a series of prototypes going from proof of concept through to the industrial design. We’re losing (have lost?) that experienced intuition in a big way…
Yeah… I’ve tried to hunt down “weird” styles of engineering to me when I can to understand them.
Chinesium is one that, as I’ve torn down a lot of things, I think qualifies as an engineering style. “What is the absolute minimum required to make this more or less work, in ideal conditions?”
The mechanical engineering of the pre-digital-age is another - I have the guts of a pushbutton mechanical door lock (buttons 1-5, you can press them in combinations, as long as each button only gets pushed once), and it took me quite a while to understand the details of how it was set/used, but it’s apparently adding machine technology. Very different solutions to a problem.
The new way of doing this toaster would probably be IR sensors, servos, etc, though since people are used to crap toasters there wouldn’t be a market unless you could make it IoT and only toast pre-approved licensed breads. 
I’d say the ‘modern’ toaster is something like the air fryer I have in the house. No longer a unitasker, not the chinesium variant so built somewhat durably, and with lots of places to take apart and maintain if you so choose (crumb tray, replaceable elements, etc)
“minimum development effort” indeed! I was recently trying to buy/set up an ESP Home node so I can do things like remotely turn my 3d printer on/off, but with multiple relays (one for octoprint, one for monitor/hdmi switch, one for the printer). In order to handle 2-8 relays, there’s a small microcontroller to drive the relays due to the limited I/O ports on the ESP8266 (8 pins total!). However, instead of making the microcontroller simple/dumb, they instead used the “turn the ESP8266 into a modem with AT commands” software stack that Espressif provides for free, and yep, their microcontroller sends AT commands to the ESP8266. Ugh, so dumb! I started to build a python script (on my laptop) to pretend “I accept AT commands”, then stopped myself, thinking “WTF am I doing?” and put the boards aside and started to design my own 2-relay boards instead. (2 layer boards, how hard could it be?)
Don’t confuse “easily fixed” with “Shouldn’t have broken in the first place.”
I’m not sure what the balance is, but if a toaster is burning out elements regularly, something is wrong, and “Oh, but they’re easy to replace!” doesn’t solve the root design issue.
lulz…
IMHO the biggest failure of a toaster (or any cooking device, really) is dealing with cooking byproducts. Crumbs, oils coming off, sooting, etc. The only solution to most of that is user maintenance.