I think Cog Sci is meant to be an academic site, thus incorporating research questions, but also helping novices with questions about scientific topics (e.g. 'how does ACT-R deal with individual differences in cognition?').
I understand the appeal of a strictly research oriented site, but I think there are some better reasons for why the site should encompass a wider range of experience in its users.
- Tags allow easy searching for research questions - research questions within cog sci will generally use much more technical tags (e.g. semantic-dementia, V1, aphasia), and following these tags should allow for you to get research questions and not noobie inquiries. Note that this isn't always the case, such as in math, where things like 'graph' and 'algebra' mean very different things to researchers and novices.
- Novices still sometimes provide insights into research questions/Researches provide insights into novice questions - research tends to breed tunnel vision, and so sometimes a novice's perspective can be invaluable.Conversely, in many cases it takes an expert to fully answer a novice's question. Say someone asks about the role of the ventral stream in vision; there is a broadly accurate answer of wikipedia, but then there is the nuanced answer of the expert.
- The University Model - Rather than going through all of the reasons for mixing research and teaching together, I'll simply point to academia. Though one point bears reiterating: by providing a place to both teach and research, you increase the pool of resources you can draw from.
That said, if the 'psychology' questions that you see in popular science magazines start popping up everywhere (I can imagine everything tagged 'consciousness' turning to crap quickly), then I'd be for a split. But I think we've should stick together until proven otherwise.