Not sure what you are trying to say, but bounce cosmology models do have predictive power.... Trying to visualize prior to big bang is not predictive power it is prior conditions.
Any scientific reconstructions of the past, whether it extends back before the start of expansion or merely back to the Jurassic, must be based on some model that is testable by what we can observe in the present. We do not look at the past, we look at its predicted effects, its imprint on the present.
We have to be able to verify early universe cosmology models using what we can observe in the present. Bounce cosmology models are no different in this respect. The bounce has a predicted imprint which we can look for in the ancient light (CMB-polarization mapping) and either find or not find. This has gotten the interest of a number of early universe phenomenologists (whose job is to think up and evaluate tests of models like this), so there are papers being written about it.
Last time I looked there were some 33 test-oriented papers on this written since beginning of 2009---they show broad agreement. Also last year there was a conference partly on this topic. It is a growing research area. If you would like to check out the papers, here's a link:
Last edited by a moderator: