I know that it is possible to detect "focus" with a consumer EEG with a single electrode and a reference point (ear clip), even without conductive gel, but is it possible to detect a flow state with any level of technology (whether it be professional or consumer)? Are flow states and focus synonymous? Is the mechanism of flow by it's nature undetectable by EEG?
Bearing in mind the fact that I can't prove a negative, I'm going to say "No, it's not (yet) possible".
Flow is rather loosely defined (e.g. "merging of action and awareness"), so coming up with hard measures is a challenge in and of itself -- even without bringing electrophysiology into the equasion. As a Positive Psychology concept, it belongs to the realm of "soft science" psychology, and this tends to make fuzzy things stay fuzzy. (Please note that I'm not using "soft science" pejoratively).
I think the best you can do is measure correlates of alertness and attention (e.g. look for desynchronization in the alpha band). If you come up with a very clever design, you might even be able to elicit ERP components related to conscious awareness of a stimulus of interest (e.g. P300b complex), but even that will be very difficult in non-laboratory conditions.
Alternatively, you could turn to machine learning to search for patterns associated with the conscious processing of a stimulus of interest. If you want to go that route, I'd strongly encourage you to read one of my colleagues' Ph.D dissertation . This dissertation proposes some novel approaches to pattern classification with EEG (and MEG/iEEG). Note that this problem is a much easier one (classification of non-communicative patients). That, along with the insane amount of noise you're likely to get and the need for a very inventive design, earns you a hearty "good luck" ;)