p.s.  It also seems like an easy thing to test.  Create a new table with a computed column on it and add a million rows of fake data.  See if it breaks.  As the old saying goes, "One good test is worth a thousand expert opinions".  😀