11-dimensional M-theory has \(\mathcal{N}=8\) supersymmetry, and \(\mathcal{N}=8\) supersymmetry is observable only at very high energies; Usually, \(\mathcal{N}=1 \) supersymmetry, or "minimal supersymmetry" is considered the most "natural" amount of supersymmetry in string theory.

The formulation you propose, would be fact exactly *equivalent *to M-theory, and would also have \(\mathcal{N}=8\) supersymmmetry. In general, compactifying on a class of manifolds known as \(G(2)\) manifolds (the 7-dimensional counterparts of Calabi-Yau manifolds) reduces the supersymmetry of M-theory to \(\mathcal{N}=1 \).

So, well, yes, you could derive a 4-dimensional theory related to M-theory by a series of T-duality relations (this is absolutely wrong, as 40227 points out above) but it's going to have \(\mathcal{N}=8\) supersymmetry, which is not exactly an ideal solution.