Do you think it was once maybe about a Ping Pong ball size or a baseball size or building size or the size of Texas? What is your estimate from non classical GR calculations and theoretical projection of rewinding the universe down to smaller and smaller size?
Glyde, it's nice of you to ask! I appreciate you asking my opinion. There was a Nasa report called WMAP5 (cosmology implications from the 5-year WMAP data) which said that in the simplest case where the U had a finite size, with 95% certainty it would be AT LEAST 10 times larger than the observable portion. (And it could just as well be 100 or 1000 times larger, the estimate was just a lower bound that it had to be at least that.)
Their number was more precise than 10. I am just speaking approximately. Their lower bound was roughly that. I can get the link to the report if you want. It's online.
Many cosmologists think of the U as spatially infinite, and therefore it would be spatially infinite at the start of expansion. And they do their calculations based on that assumption. You get approximately the same fit to the data whether you say infinite or finite-but-very-large.
So the first thing is always to remember that when people talk cosmology OBSERVABLE universe is just a small portion of the full universe that one has to model with the equations or the computer simulator. What one models is the full thing and this can be spatially infinite (even already at "bang" time) or in any case very large.
Don't confuse observable universe with the whole thing. I'm sure you know this, but people forget. It has to be made explicit to avoid confusion.
In standard cosmology, as you probably know, the universe has no edge or boundary, and matter is distributed approximately evenly throughout. So if space is infinite volume then matter must be infinite---because it is throughout all space.
===================
That is just preliminaries. Are you OK with all that?