Except that it is possible for a 50k home to beat a high land value location, even though I actually agree with you that there is probably a big 0 somewhere at the end of long cash flow stream due to the low or nonexistent land value. For example, say a home in California sells for 500k and in 20 years is worth 750k due to appreciation, but cash flowed at 3% per year. That works out to about 5.5% per year all cash (my math I know is somewhat off since i'm not really compounding, I am doing this off the cuff.) On the other hand, if you buy 10 50K homes at a 15% cap rate, and in year 20 they are effectively worth 0 (say they are fully depreciated and it costs 50k to upgrade), you still made a ~10% per year return (300% in cash flow minus 100% of equity invested). So it all boils down to the numbers: i.e. what is the rate of appreciation, how quickly will the 50k property depreciate, what is the cap rate etc. I do agree with you that people don't think like this.
I also think that related to this, cheap areas will experience a long term rise in labor costs due to out migration, further supporting the 0 at the end. Why should someone do work on the cheap when they can move to a part of country where labor can charge so dramatically more. I understand that the cost of living is higher if they move, but still the difference in labor costs seems so enormous that it still would pay to move and do for example a long commute. My two cents.