[klodolph]

Maybe we will someday, the kilogram and ampere are a bit higher on the priority list right now. Also remember that S is not a quantity that we can directly measure, so 1/T = dS/dE isn't so useful for measuring T.

The effort to redefine the Kelvin would probably use an acoustic thermometer to measure the speed of sound in a gas. This would fix the Boltzmann constant. It seems a bit silly to do this when the definition of the kilogram is expected to change, though.

In the meantime, triple-point cells are pretty convenient, much more convenient than the international prototype kilogram, or the infinite length of wire used to define the ampere. Also remember that laboratory temperature measurements tend to be much less accurate than voltage or mass measurements. ITS-90 is apparently good to within 100ppm, but 1ppm or better is no problem when you're measuring voltage, mass, or length.

Here's a presentation on thermometry, the acoustic thermometer described is what's used to accurately measure the Boltzmann constant: https://www.youtube.com/watch?v=Irr8fOLtiWc