So what your saying is that there is a discontinuity in the function that describes drag as you hit supersonic speeds? Although I agree with your statement, I'm kinda baffled on how scientists have overcome this in calculating 'drag coefficients' etc...
Understanding the assumptions behind the aerodynamics is the key to unlock this riddle. The usual equations for lift and drag assume inviscid incompressible flow. This is just a simplification to make the problem easier to solve for situations which apply that are many in aerodynamics. The reality is that air is a compressible (and viscous & not inviscid) fluid which aerodynamicists have known about for a long time.
Simply put incompressible flow reduces the math down to two dependent variables of pressure and velocity. Compressible flow however re-introduces other thermodynamic dependent variables of density, internal energy, and temperature. For incompressible flow these are
assumed constant for a fixed altitude. In reality they are not because air is compressible. The traditional continuity, momentum, and energy equations from where Bernoulli and others developed the oft used lift & drag equations from contain all these variables. However to simplify the math we make the incompressible assumptions which make the maths easier to solve because we ignore the complications due to thermodynamics.
There are host of ways all this is re-introduced when needed by aeros.
For subsonic compressible flows for instance the Prandtl-Glauert rule (known in the 1930's) provides compressibility corrections to lift and drag.
For transonic flows since the 1960's CFD was introduced to begin to accurately account for compressible flow. Prior to that empirical wind tunnel & flight test data provided the data that was curve fitted and fared to predict drag.
For supersonic linearized solutions and numerical CFD solutions exist.
It gets even more interesting for the hypersonic case!
Hope you find that interesting!