I wrote about this concept a little bit in 2008, as I saw many investigation teams (including my previous team) do all of the background and property research before an investigating on scene. This does help you paint a more thorough picture of the situation, and it may aid you in setting up experiments to capture activity, but it will also plant a seed in your mind.
I can't recall the article I read this concept from lately (though in this line of work copyright infringement seems to run rampant so I have seen it everywhere, verbatim!), but somebody higher up on the food chain wrote about the importance of doing your homework after the on-scene investigation instead of before it. Since then it has circulated to the point that it seems to be the next big thing in mainstream investigating, with I think is a good thing. If you read that a homeless man lost an arm on the property in 1932, you'll be seeing arms everywhere and hearing a homeless man talk on all you're EVPs.
While I agree that the whole group shouldn't know the details prior to the investigation, at least 1 person needs to know. I think the more efficient way to go about this issue is to have one or two people interview the client and do the property research so as to better prepare for the investigation, then simply don't divulge your findings to your team in detail. That way you can direct their activities efficiently without planting any seeds.
Needless to say, if you uncover anything that could prove to be dangerous in nature, go ahead and plant some seeds. It is always better to have tainted evidence than to have investigators in harm's way.