Personally, I don't believe that religiosity can be reliably and objectively measured. I think it is relatively easy to determine whether someone is religious or not. Beyond that I find it hard to conceive of any way to measure a persons level of belief. Religion is too subjective and amorphous to assume you have come up with universally acceptable criteria. Most people can't even agree on a basic definition of religion let alone a set of principles and traits that apply to every religious individual and group. I'm also not sure that it is necessarily all that useful. What people do with their beliefs seems far more important than what those actual beliefs are, or how those beliefs are perceived by the individual espousing them. Since "level of belief" does not seem to reliably correlate with the behaviors and actions of believers it does not seem to have that much practical use.
That said, it is a fascinating question and well worth exploring. Over past few months there have been a handful of pieces that touch on this subject. Three of those pieces that prompted me to think about this notion again are:
"Most And Least Religious States In 2013, Gallup"
"This Is the Year Liberals Take Back Religion from Conservatives"
"Understanding Why Americans Seem More Religious Than Other Western Powers"