I've been considering the asymetric nature of the puzzle setting/solving. FDev have what - two, maybe three people? - working on setting them. We have hundreds, maybe thousands, plus resources from odd corners of the internet that FDev may not even know about.
So cipher/mythological/numerical/logic problems are going to be solved very fast, as we have seen. And of course every solved problem gives us a new method to test with every new problem (an automatic reaction to morse now will be to test whether it's a drawing)
OTOH, leaving out the clues and just hiding stuff and hoping someone will run across it poses equally big problems in the other direction as we saw in the Barnacle hunt and possibly in the Formadine Rift (where we don't even know there is something to be found). The map - especially if you include planets - is just too big to be sure that discoveries will be made within a reasonable time frame (Jaques was a lucky exception). So you do need to drop clues as to where stuff can be found...
So, if you were FDev, where would you go next? I'm struck by the fact they haven't (yet) moved onto linguistics - making us search ancient human languages, or work out the sytax, vocabulary and grammar of an invented one. Nor have they (yet) created a puzzle around a genetic code, or atomic/molecular structures. But that's just two possible areas
Imagine Lord Braben has asked you to set a puzzle that the community won't solve for 'at least a week, longer if possible...' Where would you start?
