We just had the first question asking about which protocol would be best for purifying DNA. We're likely to get more questions of that kind.
Should we consider questions asking about which protocol is better on-topic? There are certain similarities with recommendation questions which are pretty much off-topic on most SE sites.
What is our stance here?
Yes, "best" protocol type questions without any sort of qualification are terrible (like this one? :P) as you can't definitively, objectively evaluate how good the answers are.
I would rather see questions of the form "How can I do X [with|without|using] Y?" (perhaps X="minipreps", Y="noxious chemicals" for the classroom miniprep question). They are inherently better questions as they will focus the discussion (such as it is on SE) on the core issue, so answers can address it and voters have specific criteria to judge them.
This shouldn't be taken to the extreme or you'll get questions that shouldn't have been asked in the first place as a much more basic, overarching question was excluded ("What kind of spark plugs should I buy so I can run my car on diesel?").
There may not be a clear answer for best - though one could do what I occasionally do on StackExchange and decide the "best" answer is the one that worked well for me, though that only applies for protocols where it's easy and straightforward to test multiple approaches.
But I think protocol related questions are definitely on topic if we're intending for this site to have any relevance to experts generally. The key will be defining them in a way that's actually answerable, but that's the key with any question.
As being one of the users with asking a protocol related question, I definitely vote for allowing such kind of questions. But I also agree with Nick T, that people who asked the questions should have some kind of expertise on the matter. Eventually, most biologists are using/developing a technology, which requires the usage of protocols. It would be nice if we have a compilation of protocols for different aims, so that people new to the field can make the right choice (objectively, because it will be based on people's votes).
People should be able to ask questions about either established protocols to accomplish a task, or as NickT points out, ask for a protocol with specific restraints or considerations. Asking for a "best" anything is subjective and lends itself better to a discussion, unless there are very clear criteria for what makes a protocol the "best." I also think that if you are going to suggest a protocol, then you should also include a general list of advantages and disadvantages to the particular method.
