The data sets analyzed included both adaptive and non-adaptive testing:

Fixed: Wisconsin SBAC, FSA, NAEP
Adaptive: California SBAC, NWEA MAP

Loosely comparing the WI and CA SBACs does suggest that a bit of the effect may be due to non-adaptive testing--but not all, or even most, of it.

Some of the usual caveats about grade equivalents (samples intended to determine normative status, rather than mastery of all grade-level topics) may still apply, though if these are truly criterion-referenced measures (designed to demonstrate mastery of standards), then the grade equivalent caveats are less significant.

...pronounced like the long vowel and first letter of the alphabet...