Generalizations are difficult to make, but you can look at the theoretical mechanisms of operation for some interesting nuggets.
In a tube, current is transferred between anode and cathode by the space charge in a vacuum. Child’s Law states that current in a vacuum is directly proportional to anode voltage (to three halves power) and inversely proportional to the distance between electrodes (squared). The speed of electrons depends solely on the applied voltage.
In semiconductors constructed of doped sandwiches of semiconductor material, Child’s Law doesn’t apply. Here we use the Mott-Gurney law. This states that the current density in a semiconductor is directly proportional to anode voltage (squared) and inversely proportional to the thickness of said material (cubed). The speed of the electrons depends on both the electron mobility of the semiconductor (assumed to be constant) and the applied voltage.
There are extra constants for calculations in either case, but notice the similarities. In both cases, we can assume distance between electrodes (whether separated by a semiconductor or a vacuum) doesn’t change. The difference here is the power for the voltage term. The generalization is that current density through a vacuum is less affected by changes in anode voltage than is current through a semiconductor.
Is this the theoretical mechanism to explain the heuristic that vacuum tubes are more linear than transistors? Maybe…at the very least it’s an interesting observation. In reality, geometry, application, and other factors matter, aside from just materials. Some tubes are more linear than others just as some transistors are more linear than others. There are bad ways to bias and operate tubes just as there are good ways to bias and operate transistors.
wauwatosa tube factory 
Since you are working hybrid you have probably noticed how many high voltage solid state devices are not built to be linear; I sure have. But then at higher voltages most transistors [bipolar or mosfets] are better off being highly nonlinear switches, for their own sake. That said- Tubes carry more inherent ‘resistance’ [it is really dynamic, so impedance is a better word for this], and are more likely to be linear as a result. By definition, too, a higher voltage supply means having a lot more room to be, linear.
There are not that many inherent electronic differences between solid state and tubes; the latter have required output transformers, with the Futterman exception noted, and this is a major difference in the sound of the amplifiers. Outside of the criminally complicated power supplies in a Futterman, there is the sonic reason his name is a curious footnote in Tube Audio History. Quasi complementary output anything is a compromise that injures music reproduction, imho.
LikeLike