Squamous epithelium is supposed to be more "resilient": your vocal cords take a beating opening and closing so much with high-velocity air, so they're lined with it. Other obvious sites of potential trauma include the vagina, skin, etc. Things generally turn squamous in the face of chronic stress.
Another thing about squamous metaplasia: Recall that vitamin A maintains cellular differentiation and that deficiency causes squamous metaplasia (e.g. Bitot spots). Dr. Sattar (Pathoma) mentioned that smoking can cause a relative vitamin A deficiency in reference to periductal mastitis in one of his lectures; I would assume the same reasoning could apply as an additional explanation to squamous metaplasia of the respiratory system in smoking.
The only reason Barrett's turns columnar instead is because it's actually columnar w/ goblet cells. This allows for a protective mucus lining reminiscent of the stomach's mucus layer. Again, the emphasis is on the goblet cells conferring an advantage to resist acidic stress. I would guess that if the cells didn't have the potential to include goblet cells in their metaplasia, they would stay as squamous cells.
