Can a new tool detect difference between two means with same effectiveness?

Hi all,

I have two gels of different crosslinking densities (gels A, n = 3, and gels B, n = 3). The gold standard test of substrate MODULUS detects muA = ~4 and muB = ~8. I invented a new engineering tool that calibrates the shear STIFFNESS of a sample (related to modulus through an analytical equation) at muA = ~0.2 and muB = ~0.4.

How can I prove that my new tool characterizes the samples as effectively as the old tool? I can calculate a T-statistic that describes the differences in the respective means (T_modulus and T_stiffness), but then is there a way to compare the T-statistics? Or some other approach I should be considering?

Thanks in advance!