What if we consider that resistance is not always a constant? Could even the semiconductor Current vs Voltage equations be written in the form I=E/R if we defined R correctly, and as a function of E, Temperature, and other factors? Why must we say "well, Ohms law doesn't apply to this device or that," when we could, perhaps, mean simply that for a particular device, R is not a constant and this Ohms law is no longer a quick, convenient equation to solve?
In other words we might not have to "throw Ohms law away" or cast dispersions because "it doesn't hold for this or that device."