So I found some info on IRLA. The big difference between IRLA and most of the other systems is that it is designed around the Common Core grade level expectations, which are qualitatively different from grade to grade, so although it has continuous scores, they are not really based on the same assessed skills from grade to grade. It also has very high mastery criteria (99% in every aspect of reading) for advancement to the next level. The routing test for determining the level to use for assessment is supposed to be vocabulary knowledge.
"The student’s baseline reading level is the highest level passage s/he is able to read comfortably, with 99-100% word recognition, expression and fluency, while answering basic comprehension (Common Core State Standard 1) and vocabulary questions (Common Core State Standard 4) appropriate to the level."
Following is a link to a technical document discussing some of the psychometric strengths and weaknesses of IRLA as a reading assessment, and quite a bit about end-user feedback. One item that jumps out at me is it has no data on standard error of measurement.
http://www.americanreading.com/documents/report-measurement-inc.pdf
