The mechanical resistance of a compound isn't trivial to define and relates very little with the strength of the chemical bonds between the atoms.
First, if you try to deduce a break stress from the bond strength, you get unrealistic values.
And second, parameters like for instance the proof stress can vary by a factor of 100 for the same alloy, depending on work hardening and on its heat treatment, despite the chemical composition stays the same.
For ceramics and semimetals, there is some hope to predict a hardness, but generally brittleness is what define their behaviour in real use, and brittleness is about impossible to predict.
So clearly, this approach is sterile. Alas, we don't have a better theory that would give numerical predictions. Nearly every data available about the "strength" of materials is experimental.