The standard picture The features in the CMB power spectrum produced by cosmic coherence (as illustrated in Fig. 12) are very special signatures of inflation (and other passive models). Similar features are predicted in the polarization anisotropy power spectrum and in the polarization-temperature cross-correlations. The absence of such features would overturn this standard picture. In light of this point, the presence of the so-called ``first Doppler peak'' in current data (Fig. 12)  reflects resounding support for the standard inflationary picture. Figure 12 also shows a cosmic defect based model [18, 19] that, through some pretty exotic physics has been manipulated to attempt the creation of a first Doppler peak. Although the broad shape is reproduced, the sharp peak and corresponding valley are not.
Figure 12. A current compilation of CMB anisotropy data with curves from an inflationary model (solid) and an active model (dashed). The sharp peak, a result of the phase coherence in the inflation model is impossible for realistic active models to produce.
Exotic variations: It has been shown that under very extreme conditions, active perturbations could mimic the features produced by coherence in the inflationary models . These ``mimic models'' achieved a similar effect by having the active sources produce impulses in the matter that are highly coherent and sharply peaked in time. Despite their efforts to mimic inflationary perturbations, these models have strikingly different signatures in the CMB polarization, and thus can be discriminated from inflationary models using future experiments. Also, there is no known physical mechanism that could produce the active sources required for the mimic models to work.
Another possibility is that the perturbations are produced by inflation but the oscillatory features in the CMB are hidden due to features introduced by a wiggly inflaton potential. It has been shown  that this sort of (extreme) effect would eventually turn up in CMB polarization and temperature-polarization cross-correlation measurements.
Another possibility is that inflation solves the flatness, homogeneity and horizon problems, but generates perturbations at an unobservably low amplitude. The perturbations then must be generated by some other (presumably ``active'') mechanism. Currently there is no good candidate active mechanism to play this role , but if there was, it would certainly generate clear signatures in the CMB polarization and temperature-polarization cross-correlation measurements.
I also note here that a time varying speed of light (VSL) has been proposed as an alternative to inflation [22, 23]. So far all known realizations of this idea have produced a highly homogeneous universe which also must be perturbed by some subsequent mechanism. The building evidence against active mechanisms of structure formation are also weakening the case for the VSL idea.