In general, it should be possible to use the plasma.  However, you need to be aware of several complexities.  For example, some proteins and/or small molecules might want to dissolve into the ficoll, so they won't be there to detect.  The phsycial interface at the ficoll might be a point where amphiphilic molecules could accumulate (with the same effect).  If some of the cells react to the ficoll, they might introduce extracellular changes -- and this is always an issue, underestimating the alterations that 'unhappy' cells can cause in the extracellular space .  And, perhaps most importantly, any handling of blood becomes a potential variable:  the time in handling after removing from the body, cells reacting to the tube environment (including, uniquely, the ficoll in this case), nature of temperature and handling conditions...  all of this will surely change the blood somehow.  It's up to you to determine the extent of those changes and whether or not they could alter your results.

If you happen to know the protein(s) that you want to detect in the plasma, you can easily do a comparative control experiment (a ficoll tube with a standard plasma tube with the same anticoagulant, drawn from the same subject).  If you are doing discovery of as yet unknown proteins, any variation in collection might cause problems.  But if you were to stick ONLY and entirely with the ficoll tubes, you might have a chance at a good research outcome.

Dear professor Craig A Gelfand,

Thank you very much for the perfect answer. This is good suggest! 

Sincerely,

Mehran