Chrome uses a graphic engine called Skia which is able to output to various devices & surfaces. It could be the screen via Quartz (Mac) or DirectX/GDI (Windows), or it could be PDF (for printing). The DOM engine (which is completely separate) knows how to call Skia to render, and then Skia outputs as requested. This leads to why the output PDFs from Chrome today are simple print-based output - because t are using a simple drawing model. However, there is work underway between Pdf Converting and Google to connect the DOM to the PDF output system so that the output PDFs can be tagged and accessible.
But for now, when using a desktop browser like Chrome or Firefox, the output is simple SVG. If you find a bug or need help, use the Google+ Community. Google Reader Beta 7 is out, see the latest release here.