Im not sure about this hardness business. It seems to me that if the object is softer than the glass it will break the glass by simply deforming it, given enough force, until it breaks. But if the object is harder than the glass it will scratch it and the crack will propagate requiring much less force. You can cut glass in a straight line by scoring it with a carbide scribe, then simply breaking it. The break will usually follow the scratch very closely. As for the question at hand, if you are inside a car and you use the blade to break the window the blade might slip off the window instead of breaking it. This could be very bad depending on where the blade goes next. I would use the handle as others have stated.
Scratches and the propagation thereof only come into play when, as you mentioned, you are scoring & cutting materials, not when the object is to simply smash the material apart. For smashing, you want the energy to propagate out through the medium from a single point as rapidly as possible without the need for prior scoring, there is no "scratch-path" for cracks to follow. Upon impact, the harder tool, or rather the tool with the most energy focused upon the smallest point, will indeed indent the glass by fracturing a small section at the point of impact... but this can actually be counter-productive as the goal is not to create a tiny hole but to eliminate the ENTIRE pane in a single blow of minimal force. Think about a high-speed bullet penetrating a pane of glass
without inducing catastrophic failure of the pane, leaving a neat little bullet-hole for forensics teams to ponder over

.
What you want is for the energy of impact to propagate well beyond the point of impact, and deflection of the pane accomplishes this. But YES, creation of a tiny flaw at the point of impact helps to generate cracks that will carry-out that propogation, which is why you want to focus the point of impact to as small an area as possible. By focusing the energy thus, you rapidly force the glass to either bend or break (to disperse the energy) at a very tiny area. With a harder material impacting, the glass will chip away more readily but will also disperse more energy thus and will not need to deflect as much to disperse the remaining energy. With a softer material impacting, the glass will not chip as readily and so must disperse the energy via deflection or deformation of the impacting material which is not an easy thing to accomplish instantaneously and so the glass ends up chipping anyway so long as the impacting material is sufficiently resistant to impact-deformation, which steel is if given a proper amount of material support.
So again, in the end it doesn't matter whether the glass-breaker is ceramic carbide or just hardened aluminum provided the energy and geometry are present. you won't be able to detect a difference in the amount of force required to accomplish the task.
And I agree again regarding NOT using the knife
blade to fracture the glass. Chopping the glass with the blade, there is less momentum in the blade and it will readily bounce off the pane, stabbing the pane risks the point slipping along the pane instead of directing its energy
into the pane, and these principles are true regardless of what the blade is made of - ceramic, steel, plastic, etc. Use the pommel, whether it is sharpened or not. If it
is sharpened, it will be easier to accomplish the task. :thumbup: