Root Cause in Analyze

I am working on a SS project right now, and what we found was that a measure within our scope is a critical X.  What we felt was that we needed to know what caused this X to vary before we could venture into Improve.  Our process is very long and complex, and after 4 weeks of continued investigating the upstream root causes of this critical X, we aren't even near explaining 95% of it.  When and how do you pull the plug on extra-scopic Analyze activities so that you can move on the measured X and make your improvement?

I am also wary of coercing my team to "cut bait" on this investigation, because I want them to be 100% in the game.  If they feel that they are bieng controlled or coerced, then they won't want to play anymore, or at least not as rigorously.

Anyone wth similar experiences?  Anyone who felt like it was time to move to Improve, but you were outnumbered?

I am remembering something I read from Jack Welch (the actual guy).  Something like, "When everyone is looking at the same information, everyone usually comes to the same conclusion."  This seems apropos, and I may have just answered my own question, but I'll post anyway in case anyone has more guidance.

Opey