This is my read of this situation: there is not any problem with IMEX or any other simulator when they warn you for "monotonically increasing values of Rs and Bo" even above your expected or measured bubble point pressure (Pb). It's completely right that for pressure above Pb, the dissolved gas ratio (Rs) and Bo are constant and decreasing respectively. However, as in ECLIPSE, IMEX input formats for PVT tables use the so-called "variable bubble point pressure option". This format requires that the saturated and undersaturated oil PVT data to be input separately with both tables covering a wide range of PVT data for different (i.e. variable) bubble point pressures to ensure that the fluid properties, and the physics within the reservoir, are properly represented when, for example, re-solubilization (by re-pressurization) effects occur after the reservoir pressure have dropped below Pb, or any other phenomenon when the bubble point pressure varies. I encourage you to read the theory on this "variable bubble point pressure option" in the Eclipse Technical Manual, and then read the IMEX user manual to fully understand how the PVT data must be input for saturated and undersaturated PVT data.