In my talk at BibleTech 2010 (kindly recorded by Weston Ruter and available at
http://vimeo.com/10489590) I talked about the disadvantages of only considering the frequency of words, and discussed some concepts like dropping the least frequent word in each verse when scoring verses, using clauses rather than verses, and also an algorithm (implemented in
https://github.com/jtauber/graded-reader for finding the "next best" verse to study next.
I cover the core of these topics in 6:16 thru 19:20, if you want to save some time
To preempt a common question I get asked: what's the status of the work I talk about... in short, I've been working the last couple of years on addressing some of the missing pieces around inflectional and derivation morphology talked about in the presentation.