I recently had a question marked as off topic about what the smallest microcontrollers are. I get that people are probably looking at it as "a shopping or buying recommendation". However, this (and numerous other questions like it) get plenty of votes and views, so they are clearly serving a need and are pretty well received. So this got me to thinking...
I didn't really mean to ask about a specific recommendation, as in "buy XYZ part from ABC company for your application". I can look up datasheets just fine, if I know what type of thing I'm looking for. I wanted to have some idea of what the approximate state of the art is, and any important issues that one needs to be aware of related to that. A perfect answer may be something like "the smallest microcontrollers use bare die packaging but need to be wire-bonded, here are some examples... slightly larger parts are available in WLCSP BGA and tend to be around 2 x 2 mm but have small pitch which requires microvias... etc etc". Specific parts are only in there as examples. Whether this came across as a "shopping recommendation" it certainly wasn't intended in that spirit.
Examples of very informative answers which are not recommending a specific part in response to a question worded as a buying recommendation question:
What are the cheapest microcontrollers? "If you want rather more than 1000 then people like Microchip have special untested supply lines where you are responsible for ensuring devices are in spec and you get accordingly low prices. "
What are some smallest, cheapest microcontrollers with USB built in? "V-USB project provides needed software for bitbanged USB and is available under GPLv2 or commercial licenses."
Which low power microcontroller for short active periods at long intervals? "I found that it was best to have 2 batteries in series to squeeze every last mAHr out of the batteries."
The meta question then is: How to ask good (useful) questions of this type? Can we reconsider whether they are on-topic, and in what cases could they be on topic?