Consider, page-25

You'll notice with that sort of configuration that the same satellites are always aligned north-south even as they cross over. Each set of five bounced up and down on a north-south line (except for the middle line which always stays on the north-south line where it intersects with the equator).

You could deploy in a single string or you could deploy in five strings only four sats long and they all go around in a group crossing back and forth. Or you could deploy in just a three strings but if that video is actually accurate then coverage will vary by latitude quite dramatically in three strings as the earth rotates beneath the sats.

My guess is that if the video is an accurate representation of how the constellation will actually be configured then they will deploy in a 5x4 cluster with each launch, slowly spreading coverage.

I seem to remember reading once that SAS would achieve full equatorial coverage with about 150 sats but I can't be bothered trying to find if my recollection is correct. If it is, then my deployment scenario is obviously wrong (which was always the case anyways; I'm not precious). It's also likely that the animation is not really reflective of the orbits.

If they instead went as five lines running in parallel "bouncing" back at each of the 15 degree latitudes then they could also provide coverage. That would be an easier network to expand since you add another parallel line and then push the latitudes out a little to cover more surface area. It would also have the advantage, I suspect, of all sats being deployed at the same speed and direction. But there are plenty of mathematicians who can figure out the best way to cover a sphere with concentric circles.

And, although an entertaining distraction, this conversation is useless. We don't really know anything about how they will fly the satellites.