Body Fat Measurement by Hand to Hand Impedance Underestimates Fat Mass

I have been preaching that for years: "Do not trust in body fat monitors". Those things will give you one of their random numbers for a six-pack of beer. Now, a study by Esco et.al. (Esco. 2010) scientifically validated my scepticism (which in part is rooted in the knowledge I acquired in the course of my university studies in physics):

[...] When compared with the DEXA [dual-energy X-ray absorptiometry], the mean values for BIA [Body Impedance analysis] were significantly lower for BF% (DEXA = 27.6 +/- 5.3%, BIA = 22.5 +/- 3.5%, p < 0.01) and higher for FFM (DEXA = 47.2 +/- 4.5 kg, BIA = 50.6 +/- 4.6 kg, p < 0.01). The results of this investigation indicate that hand-to-hand BIA significantly underestimates BF% and overestimated FFM in college-age female athletes when compared with the criterion DEXA. Practitioners should use caution when analyzing body composition with hand-held BIA in a population of athletic women.

Although the scientists restrict their conclusion to the particular population under study, the same will hold true for men and the non-athletic population as well. The charge will always take the path of least resistance. It is thus unrealistic to expect to get realistic full-body stats from measuring a current that runs from hand to hand or toe to toe (as it is the case of the above mentioned body fat scales).
As I do not expect you to have a dual-energy X-ray absorptiometer standing in your bathroom, you might as well use a cheap measuring tape to track your progress. While it will not give you a percent value it is a more reliable tool to evaluate one's dieting progress (i.e. muscle vs. fat loss) than any scale will ever be.