The last thing the universe has to be is intellectually conceivable.

In fact, for years cosmologists rejected the evolutionary cosmos as philosophically unsatisfying, that the idea of beginnings and endings and finite dimensions were simply "inconceivable". The Steady State universe as proposed by Hoyle and others seemed the natural and obvious state of the cosmos, and cosmologists had to be dragged kicking and screaming from it by hard evidence like the microwave background and isotopic abundances and the indisputable observation that very distant galaxies were different from nearby ones.

The Steady State universe satisfies the Perfect Cosmological Principle: that is, it looks the same no matter where or when we are in it, for as long or as far as we can see. It embodies no singularities, no changes of state, no evolution, it is the same forever and infinitely old and infinitely big. It requires the least assumptions: it even satisfies Occam's Razor. You only needed one a priori assumption to make it work: that hydrogen atoms were created spontaneously out of nothingness at the rate of 1/M**3/century. This would just replace the matter as it was dispersed by the expansion of the universe and it was a rate which was not likely to be experimentally verifiable one way or the other. No messy Big Bangs, inflations, singularities, or any of the other paraphernalia associated with an evolutionary cosmos. Unfortunately for the cosmologists of the mid-century, it didn't fit the data.

The real universe will be defined by the best model which fits the available data, and as new observations become available we may have to modify our views accordingly. What seems right to us should be the last thing we consider.

What I find really interesting, regardless of how this question is eventually answered, is why did the scientists of a half-century ago have no trouble philosophically accepting an infinitely old and infinitely large and unchanging universe while today's feel so uncomfortable with the concept?