I'd like to ask a question about a device that can monitor specific things. Are recommendation questions on topic?
To make this more concrete, I'd like to a weather sensing station. This would monitor things like humidity, temperature, pressure and current light level (useful for determining if it's day/night/cloudy). Would a request for such a device, to be added to a home automation system, be on topic here?
Shopping questions are off topic for Stack Exchange (including recommendations of things to get) - read the blog post for more into or the meta post for a TL;DR version.
As such, the concise answer to your question is no. If you're asking about devices or how to build them yourself and not something along the lines of "which should I get" or "which one is better", your question should be on topic.
Questions asking for recommendations of IoT devices are on-topic, obviously.
As to whether they're permitted, that's up to us to decide. Some Stack Exchange sites, such as Super User, have decided to make a blanket ban on questions asking to recommend a product. Others, such as Unix & Linux, don't see what the fuss is about.
The ban on recommendations on SU has led to the creation of Software Recommendations and Hardware Recommendations. These sites have a problem connecting askers to experts — it's especially visible with SR which has the lowest proportion of answered questions on the network. Experience shows that it's better to sort questions by topic than by question type: singling out recommendation questions was not a good idea.
So questions asking for recommendations should not be rejected just for that reason. Of course, that doesn't mean that any question asking for a recommendation is fair game — quality control still applies.
Here is a guideline when asking for a recommendation. Questions not following this guideline are likely to be too broad or primarily opinion-based, both of which are reasons to close the question.
A good recommendation question has both some hard requirements and a user story. The hard requirements are constraints (on price, connectivity, power consumption, processing power, storage capacity, etc.) that must be met. The user story allows ranking between candidates that meet the hard requirements.
See also the question and answer FAQ on Hardware Recommendations, and the question and answer FAQ on Software Recommendations.
I think if you're talking about the sensor technology in isolation, Raspberry-pi or Electronic Engineering sites are more suitable. If the question relates to network specific considerations, crowd-sourced data mining or calibration, etc - only then there is an IoT specific angle.
If its about how to extend an existing home-automation system (with IoT specific parameters), then it is more clearly on-topic. Feels like there is some grey-area at one end of the scale.
To elaborate:
Questions about finding an SBC with a given feature set, or other 'off the shelf' components don't seem like particularly useful questions, and seem to often be asked without much thought. I think we should discourage these questions, and maybe direct people to a different site instead. Alternatively, encourage them to ask about how to solve their own problem, rather than asking directly for a solution.
Questions about if something exists, or emerging technology which is specific to IoT may be on-topic, because there might be some deeper explanation to address the problem of 'why can't I find this thing which seemed in vogue last year'.
Questions about search terms and approaches to selecting devices seem appropriate because they are about improving the search apace, and teaching methodologies.
Regardless of whether or not it's acceptable on other SE, I strongly suggest not to allow this here. The internet of things in general is a very fast changing field and suggested recommendations are much more likely to have a shorter half life. Most recently during the private beta, pebble will no any devices and this won't be an isolated incident in this field.
