You could call this a "smoldering hot topic" in sputtering - everybody asks for it at some point but the answers are often disappointing.
In principle, you can measure the formal temperature of a plasma by simple pyrometry and pretend that that's the temperature everything has. Of course you have to check with OES references that none of the characteristic wavelengths in the plasma coincides with the wavelengths your pyrometer uses.
Then you have a temperature which you can write on a report and pretend to have done it. However, the energy distributions in sputtering plasmas may be highly inhomogenous, especially when you come close to a biased substrate. You could even debate if the states inside there are homogenous enough so that the mere concept of a temperature makes sense in there.
There are concepts for calculating the energy distributions of particles impinging on the substrate from the plasma, but everything I've seen so far had major caveats on it.