Several things going on here. Of course the ceramic glass transmits some heat. If it didn't, there would be no need for the IR coating some have. Or you can just look up the properties and see it does transmit quite a lot through the IR spectrum.
(broken link removed to http://www.ceramicglass.co.uk/downloads/neoceram_charts_1.pdf)
The glass is also heated by the fire, so it is radiating as well, and it's very rare to have one mode of heat transfer without the others so there has to be convection and conduction, too. But what of the heat which is reflected back into the firebox? - it can't just disappear. Part of that is absorbed by the firebox which makes it hotter - which leads to a cleaner, more efficient burn - generating even more heat. The hotter firebox raises the overall temperature of the stove - which means more heat coming out the top, sides, etc. In short, what heat doesn't come through the glass leads to a more efficient burn and will come out of the stove as heat from other surfaces.
As for dirty glass, the IR coating is generally a metal film which is highly reflective in the IR spectrum, though good transmission in the visible. Carbon soot is the opposite...good absorbance in the IR and visible range. Hence, it's most likely defeating the purpose of the IR, coating - letting the glass get a little hotter, radiating a bit more heat to the room, but letting the firebox get a little cooler and less efficient.