I wonder if this solves the problem: If listing breakpoints from small to large, specify min-width but omit max-width. I saw this shown (not recommended but tangential to another point) at <https://zellwk.com/blog/responsive-typography/>, as accessed 2-18-17. I suppose if listing breakpoints from large to small, specify max-width but omit min-width. This would seem to yield a range from the first value to any value less than the second value even if fractional. Or would browsers be inconsistent in dealing with this? Like, would the omission of the second -width value cause an overlap in some browsers? Should the spec be more specific in supporting this solution to encourage inconsistent user agents to act mathematically consistently?
--
Nick