#1. All the other versions mess with the layout morphing the rock, and makes it harder to distinguish the tiles properly (and results in a lot of additional work for both Toady and the artists).
@[HYBRID BEING]: There are no diagonal minecart tracks in DF. Introducing such tracks would require a fair bit of changes to the code and memory structures, and I believe a crucial structure is already out of bits. Thus, that's a technical game change, not a display issue.
Why should paid artists not due to deliver a product for at least a year worry about the "extra work" of drawing diagonal tracks and corridors? Is there some upcoming deadline only you know about?
Everything else is being modified by Toady and the artists as they put these sets together, why not minecart tracks? Isn't the point to make a fine looking game for the premium release (within the limits of flexible procedurally generated imagery)?
Also, could you post a picture showing how it makes it hard to distinguish the tiles, not sure I understand the logic (apologies if it's obvious, not all us are pixel artists).
I'm about as far away from an artist as you can get... Maybe you find the posted examples completely clear, but I do not.
The artists should (and do, as far as I've seen) care about delivering an acceptable level of quality for the full range of assets that are to be included at release, and ought to advice Toady on where to best spend their time from an artistic perspective. Toady obviously has the final say, but it should be part of the artist role to provide advice within their areas of expertise.
The deadlines "only I know about" are, as voliol mentioned, Toady's stated intention, and what (only I, apparently) believe is a contractual obligation to deliver before a two year deadline from the commercial version contract (landing the time somewhere in the vicinity of mid January to end of February). At time of delivery, all contents obviously has to reach a minimal acceptable level of quality (which has been mentioned before in this thread, I believe), although that does not mean work will stop at the first release: there are indications, I believe, that work will continue for an unspecified time after release during which nice additions (as well as pure improvements) could be introduced. But yes, I believe there's a lot of work and less than half a year in which to do it (assuming there will be some closed beta testing to avoid some of the usual crash bugs at release, but testing is the first thing cut when deadlines are not met...).
Any work Toady spends on adding new game play functionality for nicer graphics has to be weighed against time spent on stress balancing, UI overhaul, logic overhaul (an extension of the UI work, such as e.g. the military UI which probably requires functionality changes underneath as well, stockpile refinement [quality/wear/size] and re-partitioning, etc.), the rest of the graphics, and crash bug fixes (and I'd disable Villains by default in the Premium release rather than release that unbalanced mess as a standard feature, or it will hurt the commercial game reception badly). Is diagonal track support more important than some of these things (many of which will probably have to be cut down in scope due to a lack of time)? Not in my opinion.