This suggests that hypothyroidism shortens the circadian cycle, while hyperthyroidism does nothing. This is disappointing, and also counter to what I expected. However, it could still be consistent with my expectations if these effects are behaviorally mediated. That is to say, it could be greater stimulation that keeps them awake longer (which in turn effects the body clock). If that is the case, the alertness/dullness effects could be countered by depressants/stimulants, which would reveal the direct circadian impact.
Additionally, I'd like to know if T3 doses at wake time could provide any entrainment (or at night time). The half life of T3 is short enough that reveal phase effects. T4 only demonstrates net effects; like getting light exposure all day long, T4 may both delay and advance circadian phase.
Also note that all conditions have a distinct rightward tilt when graphed. That is to say, they have a body clock longer than 24 hours. It's quite plausible that the thyroid hormone could interact with zeitgebers.