I imagine a power density of 5-10 W/mm^2 would be adequate for foam, at least for CO2.
I ran through some quick calculations, but I'm wondering if I missed a decimal place (or five). A DVD laser diode has a wavelength of around 640nm... good optics might get you to a spot size of around 10-15um. My in-the-head calcs say this is a beam density of 390MW (yes, that's mega!) at that spot size... the advantage of this semiconductor laser over a typical CO2 laser is the wavelength, which allows you to focus to a smaller dot size by a factor of 1,000 (and hence have a higher power density by a factor of 1,000^2).
Can you achieve that? I don't know, never tried, but I imagine the focal point is going to have a very short length. I'm merely looking at it from a math standpoint, and I have no idea if the power will translate in this way, particularly at a different wavelength than CO2.