Ultimately I suspect Arduino boards gathering dust on the self are the overwhelming majority. People are either interested in it and rapidly move on to more specialized applications (Hobbyist multi-rotor electronics quickly graduated from 8bit AVRs to 32 bit ARMs) or they aren't and the dev kits find room in the closet.
I figured I'd either learn android-development, or hardware-stuff. Bought an Arduino kit and never looked back. In my limited experience people dabble at such things then move on, or give it up entirely there isn't too much of a middle-ground.
I still dabble with ESP8266 things two years later; most recently I hooked up a simple 433Mhz radio-receiver to a board, and wrote some code to receive/understand/decode the radio-broadcasts a cheap off-the-shelf wireless temperature/humidity-sensor transmits. This allows me to log the temperature/humidity of our sauna.
Usually I'd wire up a sensor of my own, but I figured it would be less-risky to use an off-the-shelf sensor/transmitter. The idea of mixing steamy saunas with a USB-PSU was a dangerous one, and I even balked at using a rechargeable battery.
There aren't many things I buy for my hobby/practice, just random sensors, or accessories. Most of the time I just order them from AliExpress and take the time-hit waiting for them to arrive.
It's all fun stuff and of some usefulness to me personally, but I don't think much of anything I've done has any professional applicability; other than generalized familiarity with electronics and components I wouldn't feel comfortable producing a real product.
Are there actually consumer electronics out there that are just arduino examples glued together?
A lot of the existing power-switches that can be toggled with an app / radio-button are little more than an ESP8266 and a relay. The sonoff-line of smart-switches would be the obvious example.
Sonoff also make and sell RF<->WiFi bridges, but I guess the target market is already hackers rather than consumers for things like that.
Otherwise I've seen a few wifi-controllable LED-strips, and similar things that I'm pretty sure are identical in terms of hardware to what I've built. You can see some crowdfunded projects are just scaled-up home-toys too.
I'm struggling to think of better examples, but I'm sure they exist! In terms of product though it's interesting to evolve something from a bundle of junk loosely coupled to being a thing in a pretty case which can be enjoyed by others. I setup a trivial display to show tram-departures from outside my house, which the whole family uses now and loves. It went from a cardboard-box with a pile of stuff in it to a 3d-printed display, and then later I added the ability for it to be controlled via a HTTP-server, running as an Access-Point when it was initially setup. The whole thing became very very user-friendly. (But at the same time the only real user is ourselves, and trying to sell it would be a bit difficult.)
The Arduino IDE is also compatible with various other microcontrollers, including the STM32 family which are extremely widely used in production electronics. Typically their production code isn't written with the Arduino tools but you can most certainly prototype with it.
Some of the newer ones can trace their lineage pretty directly to Arduino clones running TMK/QMK firmwares. The commercial version is usually going to cut out as much of the Arduino bits as possible and go for a soldered on microcontroller once you're building for sale.
You'll find AVR chips controlling many, many devices in the wild. My Lulzbot 3d printer uses an ATMega for instance.
First I became a local volunteer for ultramarathons, health care emergency exercises, things like that.
Next I developed a bunch of communications standards to support people in wildfire-type emergencies.
Then I just got really involved in learning about radio tech that uses the internet, and started working on a webSDR project to get a local server set up.
In between all of that I've participated in some ARISS SSTV events via the International Space Station and have some nifty commemoration certificates on my wall.
And finally I made a bunch of new friends and became part of a group that actually mostly observes a really high standard of communications decency and loves to support learning and technology. Through the community I've befriended astronomers, cops, engineers, firemen, physicists, all of those things I idolized as a kid.
This is to say nothing of the gadgetry, the ability to reach out across the world using any number of methods. Randomly getting in touch with a Japanese radio operator whose daughter lives just down the street from where I used to live in Japan. Talking to a Hugh from Ireland and telling him about my son who is also named Hugh. Getting my kids on the air and asking a bunch of radio operators across the US to take turns wishing them a happy birthday just for fun. Knowing they're enjoying it too. Fun little moments like that.
Adafruit has been driving a lot of usable, rewarding electronics. They've been pushing LEDS (mostly WS2812 and APA102) in usable, prepackaged form factors. I really love their LED strips that are already diffused. They're beautiful and immediately usable. They're a little expensive, but I use their products to try out my ideas. But besides them there's not a lot of companies who make physical hardware appealing and available.
Side-note, why aren't there LED strips with a 3.3v logic level??? Almost all microcontrollers have a 3.3v logic level but APA102/WS2812 have a 5v logic level. Very annoying to convert.
A lot of micro-controllers are run from batteries, especially high-volume products, so there’s always going to be a pull towards power efficient languages, even if your app has wall power available.
Second heeen’s comment on using 3V3 directly. I built Halloween costumes that ran directly off an 18650 powering the Arduino and driving the APA102 LED strings. No issues as the batteries ran from 4.2+ to 3.7 or less volts. If it was life safety, I’d care, but for blinkenlights, I’m willing to push outside the datasheet.
Specifically check out bibliopixel for controlling LEDs in python. https://maniacallabs.github.io/BiblioPixel/
I hope companies like osh park, sparkfun and so on never go the way of radio shack.
That's not to say hardware stuff can't thrive, just perhaps not at the scale venture capital would require.
The counterpoint to this would be: if your child built an app and showed it to their friends, would they be impressed? How about if they built a smartphone?
IMHO, the quick pace at which everything change is a major cause of this complexity. Even for professional products, for example it is already difficult to find hardware supported by redhat 6. As opposite, see the effect of stability of arduino or ne555.
I see them as a hobby for electronics people, non-programmer computer types, or just for fun for anyone. But given a choice between some cute electronic that won't leave my house and building a web or github based side project that thousands of people could potentially use, I'd go with the latter every time.
I say that after giving up on my own gaming electronics thing I was building until I realized my time could be better spent and I could buy a finished one from China for half the price.
Rather sad to read that electronics is thought of as a toy. It is the "body" which software runs on. I agree cute electronics is probably not that useful but the are many others.
For hobby projects (both software and hardware) I generally don't care if they leave my house or if one or thousands see them, that's something I do for me, not to fulfill some hypothetical need of the world.