Why didn't 3D printing become a standard commodity after around 2012?

Not sure an anecdote will generalise to all applications, but certainly in research we very often resort to 3D printing as a first pass and especially when we need things fast, cheap, when its utility does not justify producing engineering drawings, or when parts are too complex for more conventional subtractive fabrication methods like CNC machining, i.e. taking material away rather than adding it on.

Startup cost: You can get decent plastic printers from £800+ to a few £k now that can print ABS-like materials, e.g. Ultimaker, Makerbot, etc. More expensive MJP-type printers like the Objet and Projet still go for tens of thousands. Like DKNguyen said, metal printers require specialised training to use as it's pretty dangerous and expensive. Clearing the metal supports is difficult - imagine breaking thousands of tiny metal struts. If you're talking about it being a standard commodity like a laptop, then the plastic printers above are your best bet.

Material strength: ABS-like stuff is pretty strong at room temperature - Lego makes its bricks out of ABS. Material properties are published by manufacturers so you can compare the range of filaments out there.

Material cost: A 1kg roll of filament for the Ultimaker goes for about £30-£40. Support costs a bit less I think. You could probably print a new printer housing with this much material.

You could go full-out MacGyver and build your own for much cheaper. There are loads of tutorials out there now (as opposed to 2012 I'm guessing) that teach you how to build your own 3D printer.

On the opposite end of the spectrum, you now have gas turbine manufacturers testing 3D printed components, e.g. blades and fuel nozzles

I guess the main reasons why 3D printing is not in larger commercial use are quite varied depending on the industry. Take Lego bricks for example, these are made by injection moulding which can produce many bricks from a single mould very quickly and to a very high degree of accuracy and repeatability. If you tried to print them even with multiple print heads running, they'd take tens of minutes depending on print speed and layer thickness, and the surface finish would be very poor. You'd need to hire many people or buy additional machinery to break/dissolve the support material and sand the parts so that they have that nice polished look. The cost to switchover from more established and well understood technologies would be enormous and may only be justifiable for very high cost parts

If you wanted it for home use, it's cheaper and more convenient than ever before. Want a computer case fan? Print it. Too lazy to hold your phone while watching videos? Print a holder. Need a waste paper basket? Print it. Just as long as you're not planning to sell the printed parts.

One facet of this discussion which hasn't yet been discussed is 3D modelling literacy.

Think for a moment about regular '2D Printers'. You may have one at home - but what do you use it for? For printing out artwork to put on your walls? Glossy photo-books for your coffee table? Magazines? Cardboard Packaging? No - mostly, you will be printing items that you have created yourself, such as letters or posters perhaps. Users recognise the limitations of home (2D) printers, and do not expect them to produce 'production-quality' artworks. Their usefulness is in their ability to provide custom items, with no delay. The primary use for home printers, is to print user-created content

So, too, it is with home 3D printing. As discussed above, home printers are no good for simply replacing commercial solid items such as car parts, guns, etc., and we should not expect them to be, in the same way you don't expect an inkjet to print out wallpaper. What they are invaluable for, however, is user-created content, where the object in question cannot be purchased. I've lost track of how many 3D printed 'bits' there are around my home, but, over 90% of these are items that have been designed by me, to fit a specific need.

The learning curve for 3D modelling, is much steeper than for text-editing. Users can open MS Word, type something out, and press the print button with no prior experience at all - sure the finer details of formatting can take a while to get to grips with, but that first hurdle to entry is really low. 3D modelling is much more difficult for a beginner.

I see all the time hobbyists with their new 3D printers who are primarily printing designs that they've downloaded from the internet, to varying degrees of success. This is like only ever printing out photos that you downloaded from google images on your inkjet.

So - to get back to your question, the 2012 hype surrounding 3D printers was linked iirc to expiring Stratasys patents on FDM/FFF technology, and this has led to much more readily available consumer printers, but these are mostly pushed by smaller companies and open-source projects. Why haven't any big corporations got in on the act? Why is HP targeting the high end industrial market only? I imagine they've done their market research, and discovered that there simply isn't the demand. Users who are unable to create content, don't have a need for a printer.

Kids at school right now are seeing 3D modelling being introduced as part of the curriculum, in a way I remember being taught about MS Word etc. when I was at school - I'm sure it won't be long until most adults have some basic 3D modelling capability, and then, perhaps consumer 3D printing will 'take off'.

Most people when asked want to do quite big items. Why wouldn't they? After all they live in a world where objects are table and chair sized. 3D printing usually can not do objects much larger than a pocketbook. Which is in the lower end of the size of things that people want to do. I mean sure at work we have a printer with a cubic meter envelope. But even then people usually back away from the project when they realize that their print would take ten days to complete under optimal circumstances.

Second most people want to do everything in one session. They dont want to design piece 1 with 3D printer piece 2 by tablesaw and go to hardware store for other parts. Its too convoluted, too far away from the goal and takes too much time. People usually have a short window of interest during which they need to be able to get their thing done or it will just never get done because other things fight for your time.

Third most people dont see inside of objects thy see the user interface and casing and want to be able to do those things. So having the slightly rough look and feel is a turnoff. People are shocked when they realize that the print has to be cleaned and prepped for final quality. While you can do very nice final surface quality with high end printers that cost tens of thousands its really expensive to use. People often get shocked when the consumables of their high end mockup print is in the hundreds.

But yes 3D printing is very useful for making jigs, fixtures, pressbrake dies, small replacement parts, moulds etc... Just not something most people do. And from manufacturing point of view its too expensive to print even relatively short batch runs. Because the cost and time of 3D printing is linear ( O(N)) while many other processes the cost is lower but also less than linear (say log(n) or even lower for something like injection moulding) since their setup cost can be high but after that each part comes out at minutes or seconds at costs of cents. This makes it easier to justify the design cost. Very few products would be viable if you needed to pay for design of each.

So there is no big need for mass customisation of hardly structural, small, nonvisual parts. And those industries that need these are still immature wait another 10 or 20 years. But maybe your spare parts are allready printed as keeping a back catalogue is expensive, why not just print one. Car companies allready do this to some extent.