I think there's a parallel to hacking though. On our weekend projects, our babies, our github projects we're most proud of, we like to make sure that we get it right: clean code, beautiful abstractions, efficient implementation, etc -- because we, as hackers, get a kick out of that. We have a gut feeling what is the right way to write a certain program. Like a painter knows where to draw the next line.

...but does that mean that we do the same in our day to day jobs? The code that you have to fire out to meet next Friday's deadline? The program that will make it into the next product that needs to hit the market now? I'd bet almost all of us know the feeling of, "ah, if only I had a little more time, I know exactly how to code this the way it should be". But this is business so (a) the time you need just isn't there and (b) what the heck, you don't care as much about the accounting app your company sells anyway.

So, yes, I'm sure that ebook publishers aren't idiots. And they know that ebooks can be done differently from printed books. But they are businessmen, too, and so the question is always: will doing these changes give them more profit? That is, will the additional costs of e.g. moving the glossy-paper pictures from a central position directly to the parts of the books that refer to them result in more sales? Or sales at a higher price?

I figure if the answer were yes, the publishers would do it. But as long as they get away with sloppy editing and one-to-one copies of paper prints, why would they?

The additional costs are not material costs, but that people have to sit down and make these edits. And that's not just one guy spending half a day to do that.