Does it help for me to call you a jackass? If not, you could try not calling it vanity when you admit that they're trying to emulate the original and failing due to lack of knowledge. It's not helping anyone understand what's wrong with these files, nor does it encourage anyone to work with you.
These issues and their solutions and workarounds have been discussed on this forum and the DP formatting forums multiple times over multiple years without forward progress. What would it take to actually make forward progress? It's a serious question. One could put together a design book of suggested solutions, but guess what DP has already done that and most of their suggestions don't work, if they had tried them. There are also simple techniques that work wonderfully, but people refuse use them, here's a couple examples: 1) Don't use bottom margins on paragraphs. Only use top margins. 2) Don't indent paragraphs AND put a line space between paragraphs. Choose to wear either belt or suspenders. Don't wear both. 3) Don't try to emulate drop caps. HTML doesn't support drop caps, which means your attempt to emulated them will fail. Raised initial letters, however, can be easily accomplished--if you insist [I personally think they look goofy with modern fonts on modern computers.] 4) It is OK to default some or all of the CSS. Most devices will make reasonable display choices by default. When a formatter chooses to default a style, that IS a design choice -- PG epubmaker should not override that default. Doing so may break a deliberate design choice. Conversely attempting to specify all styles aka "CSS Reset" will not work on many devices. 5) Don't optimize for one platform at the expense of breaking things on another platform. PG has the online epubmaker. There is ADE, and desktop NOOK, and Kindle Previewer, etc. It is not hard to check your code and see if it will work or not. Guess what, if you do check it out, you will see that it fails, and then maybe you will choose to fix it. Why not then fix it? Again, I claim it IS "vanity" when people try to "over-engineer" a simple problem. Any one has worked in engineering for any length of time sees over and over again this tendency to engineer "baroque" designs with doohickeys and filigrees everywhere until top-heavy, the whole thing tips over and sinks ala the Vasa. How about working instead in the opposite direction? Make a design which is as simple and clean as possible, while capturing the essence of that which the original author was attempting to communicate. I would also point out that the goal SHOULD NOT be to try to exactly emulated the original. If you want to do that, use page image digitizations aka photocopies, or use a PDF or TeX format. We are trying to capture the essence that which the original *author* was attempting to communicate. We are *not* attempting to capture all of the design choices of the original *typesetter* -- which were often not that great design choices in the first place. My suggestion, again, is very simple: When people submit HTML code which fails on one or the other of the major platforms, then other volunteers should be allowed to fix that broken code. It is NOT "vanity" nor is it "snowflakes" to want PG to publish code that works on most major platforms. Rather that is PG's charter! "Write Once Read Everywhere." But we are not allowed to do this. Why? A: PG is afraid to hurt the personal "vanity" of the original HTML submitter. How sad.