I've since learned in this thread that Chrome and Firefox are not as good as Safari for this technique - it hasn't impacted me much since I only use Chrome/Fire

I've since learned in this thread that Chrome and Firefox are not as good as Safari for this technique - it hasn't impacted me much since I only use Chrome/Firefox for development, mostly. And although I do occasionally check the produced PDF's, the layout doesn't matter to me at all since I use a cmd-line grep or combination of 'pdftotext' to find the page, open the PDF, and click the link to go to the original web page if I need to .. haven't found a single dud PDF in the collection in a randomised sampling, but then again in 20,000+ files, there's bound to be one that didn't make it through the rendering pipeline, but so far, hasn't been an issue.