Most people working in the field which is sometimes referred to as `CMB phenomenology' are currently struggling with the same question, in one form or another (e.g. ): how does one confront the concept of Inflation with the concept conventionally known as Proof?
It seems clear that we are now in a position to say something beyond the Big Bang paradigm. The CMB led us to accept that the Universe used to be hotter and denser, and more recently to the conclusion that structure built up through gravitational instability. Now it appears that we are learning something further, something about the origin of the perturbations themselves. But just exactly what that next step is, and how to phrase it, is altogether less clear. For lack of anything better, let me phrase my own current belief, which I challenge anyone to disagree with:
Of course the interpretation of this statement depends on the exact definition of the two crucial words. I'm sure that I don't know what I mean by `proof'. But by `inflation' I mean some mechanism which gave rise to a roughly scale-invariant spectrum of adiabatic perturbations, over a wide range of scales, including those which are apparently acausal. The only causal way we know of to do this is to have the scale factor accelerate ( > 0) at some time in the early history of the Universe . And we can argue about whether something that achieves the same end result is just isomorphic to inflation, even if interpretted differently. `Inflation' does not necessarily carry with it the extra baggage of an inflaton potential etc. - although hopefully the connection with particle physics would follow later.
It used to be that discussions of inflation focussed on the number of e-foldings required to solve horizon, flatness, entropy and monopole problems. However, at the present time the paramount concern is making those darned density perturbations. And inflation gives you a mechanism to do that, for free! It appears that we are learning that the Universe has inflation-like `initial conditions'. Time will tell whether that means that the Universe was once dominated by some vacuum energy density, and whether we can learn details about particle physics at ultra-high energies. The promise of the CMB is that it provides a way of probing density perturbations while they were still in the linear regime (i.e. simple). Thus we may be able to learn details of how the perturbations were generated which may lead to direct information about physics at energies at the GUT scale, or even the Planck scale.