There are actually two sets of graphs. You are correct that the first set only measures thrust as a function of kicking frequency, but the second set of graphs measures thrust as a function of the toque delivered to a kicking leg (simulating muscle exertion) that is generating the thrust. It seems reasonable to assume that VO2 consumption would be proportional to muscle exertion.
As stated at
Efficiency & Geometry — Truefin
<Testing VdotO2 is rather involved, and as Pendergast noted, VdotO2 may be effected by divers "consciously or unconsciously" altering "their ventilation independently of their VdotO2". For this reason, the Truefin machine tests correlating input power with both kicking frequency and fin thrust may be considered objective tests while omitting human interaction.>
From:
https://www.researchgate.net/publication/10677199_Evaluation_of_fins_used_in_underwater_swimming
< The physics of underwater swimming with fins is complicated, as demonstrated by the
data from the present study showing fins with very different designs can have similar energy
costs of swimming (Attack vs Apollo). Some fin comparisons are reported for swimming at
slow speeds with measurements of air consumption, however due to the low ventilation, the
diver can consciously or unconsciously alter their ventilation independently of their VdotO2.
This invalidates these types of studies. This point was emphasized in this study as the reliability
of ventilation and velocity was very low (r = 0.54), while the reliability of VdotO2 and velocity
was very high (r = 0.94). Fins designed on the basis of physical principles for airfoils or
propellers do no necessarily lower the energy cost of swimming (Apollo vs Appolo taped). This
demonstrates that, based on our current understanding of the physics of underwater swimming
with fins, theoretical models have to be evaluated empirically as was done in this study. >
One thing I am not clear on though is that in the Pendergast paper, the O2 and CO2 fractions in the expired gas were determined using a calibrated mass spectrometer, so why does it matter if the diver alters their ventilation, they are still measuring O2 consumption. But evidently at slow speeds VO2 tests are unreliable.
Joe Maresh