Actually, MAX can be used in this case to eliminate NULL values, as well; since NULL represents unknown data, using MAX eliminates NULL, as the MAX is either a NULL if the column is only NULL, or actual text data if it's anything that isn't NULL:

SELECT 1,userid,MAX(emailaddress),MAX(StreetAddress)

FROM @MissingTextData

GROUP BY UserID

This will handle things in your test case, but again, if there's variance in the data (like a single UserID having multiple email or street addresses), this won't work. If that's the case, a more fluid solution can be developed if expanded sample data is provided 🙂