Will the increasing population of the southwest (which currently appears to be a function of both magnitude and distribution, to use your apt words) actually stabilize and decrease as the demographic transition model theorizes, or will we simply have ever-expanding shantytowns like the one growing in Fresno? So let's take these in turn:
Although the homelessness problem is chronic to some degree, clearly our current economic problems have worsened it. And in some California cities such as Fresno, apparently the problem is growing because people who came to work in an agricultural industry that can no longer support them -- due to drought-related cutbacks in agricultural production -- are nonetheless remaining to subsist in shantytowns.
Seems to me that the panoply of current and potential issues and problems -- both in the U.S. southwest and elsewhere around the globe -- are far more complex than the demographic transition model can account for; and, of course, the model doesn't even attempt to address migration issues, which are eventually bound to become an enormous problem world-wide due to climate change.
Another question (I'm guessing I know what your answer will be but...): Do you think there should be any firm government-imposed restrictions on further development in the southwest (e.g. no more new housing projects) due to anticipated decreases in the amount of available freshwater throughout the region? 
- The demographic transition takes awhile to manifest, and it takes place at a certain income level. Given that the US is well into the transition (low birthrates, low death rates), what we are really seeing is "misplaced" population. In Fresno, people are in the "wrong" places because they have lost their jobs and cannot afford to move.
I am not sure about the solution to that problem. Some people advocate one-way bus tickets; others think that people should be given loans to move; still others think that employers may pay for people to arrive in places. The first two ideas make no sense if there are no jobs in the area. The last one will probably not work because it requires that employers know where "cheap" workers are located. Besides one obvious example (in Mexico), there's little to do about that.
I'd prefer to support people via unemployment payments/welfare that are high enough to stay alive, but low enough to encourage job search.
- Government is already restricting development (via "show me the water" legislation that requires new developments to have 20 year supplies). Although I am free market person (higher prices would have minimized sprawl), I prefer regulations on growth per se to regulations on water that are designed to slow growth. In this post, I describe how "cap and develop" could work.