This follows on something I raised yesterday (see 'class or id selectors with javascript').
I've done a template for a set of "location" pages which will be aimed at and optimized for people who are looking for particular goods/services in particular locations. These pages will be linked to an order engine on a separate website so that -- for instance -- if you selected Product A from newsite.com/reading.html, you would be transported to an order form on existingsite.com on which the product and location were already selected.

And what has this to do with CSS, you ask? Bear with me, please.

In establishing design criteria for the new site, one of the major appeals of CSS layout (as opposed to tables) was the prospect of leaner code in which content was more visible and accessible to search engine spiders. So in my mind, anything which would bloat the code is at least suspect and must be justified.

Which brings me to javascript.
I am persuaded that we need some; at very least, to make the links to the appropriate dynamic pages on existingsite.com
It has been proposed that we also use javascript as an internal work-saver on newsite.com to handle images which are repeated on many pages. That way -- the logic goes -- we can change an image on many pages with a single change in a .js file.

I'm wondering if the same thing could be accomplished (with far fewer visible lines of code in the html) using the background-image property in CSS.
If the image name and location can be specified in the rule for a DIV ID or DIV CLASS -- let's say: .prodimage1 { (other rules), background-image: url(../images/prodimage1.jpg); } -- then that class (or ID) could be used on many location pages, and the image on all of them could be changed by making a single alteration back in the CSS file. Or am I missing something?
Assuming that this idea would work, can anyone point to disadvantages as compared to solving the problem with javascript?
Thanks, Cam