It is difficult to predict but I would imagine the slope of the isolation and insertion loss graphs would continue in the same direction. I don't think it would work at all at DC so somewhere between DC and 5MHz it will simply cease to work. For low frequencies you would probably be better to use a simple resistive divider.
You can compare with the data provided by MiniRF for a similar splitter (5-1600MHz), but for which, the manufacturer provides information below the minimum published frequency.