As discussed many times previously, there can at times be a severe lack of 'non-homework' questions here on chem.SE, meaning that it can be very difficult to actually engage (which has, I admit, made me grow bored of the site at several periods since I joined).
Several things have been doing in the past to try and solve this, namely the bounty dance, aiming to resurrect good quality questions that never received the answers they deserved.
Informally, its been suggested in the past that regular users should make more of an effort to ask rather than just answer, and indeed this does happen to a certain extent (I've certainly been making an effort since I've noticed the lack of new questions here), but there could potentially be a more formal solution to this in which we actively try to make more good-quality questions.
There was a post on meta.SE about 'weekly topic challenges' to encourage users to post high-quality questions.
The essential premise is that we have a meta post where users can suggest a tag that they think there should (or could) be more good quality questions about.
These then get up/down-voted, and every so often (at some regular interval), one of the mods here on chem.SE initiates a featured meta post inviting users to think up and post their questions related to the given tag.
With good quality questions, hopefully, we end up with good quality answers (and any bad questions will necessarily end up down-voted/closed). There could also be some mechanism of incorporating the bounty dance so that the 'best' of the questions has a bounty applied, to further encourage people (not that its needed...).
This has been happening on several sites across the network, often with good results (graphic-design.SE is a good example of a relatively low-throughput site doing this to good effect, see the linked meta post here for their discussions).
Aside from the obvious benefit of getting more high-quality questions, it could potentially also help us define the scope of tags better. This kind of challenge could be used to expand the number of questions for tags that we deem useful but that are currently under-represented, as a start, and purely selfishly, i'd suggest pericyclic (theres enough possible questions and enough people around here to answer them.
I think one crucial thing here is to not make a load of awful Q&A because nobody knows anything about a very narrow topic, i.e. asymmetric-ruthenium-photoredox-chemistry-in-aqueous-solution.
Some other possibilities that could benefit from more questions: supramolecular-chemistry, stereoelectronics, x-ray-diffraction.