Paul,
I don't disagree with you on a lot of things. It's just when I agree and don't have anything to add I don't say anything.
There have been some promoter studies that have done comparisons of promotor regions of specific genes, but the large scale comparisons are just going to be possible soon due to the sequenceing of briggsae for elegans, mouse for human, and psuedoobscura for melanogaster.
In terms of sites that recruit enzymes which modify histones and such, much less has been done as the sites aren't very well characterized yet and few have been experimentally characterized so it is hard to know what components are required.
One of the main problems with studying this kind of stuff is the density of sequence needed over evolutionary distances. Since it is much more variable, and indeed needs to be so to suit the different organisms uses of genes, you see very little in highly diverged organisms. Also since the functional aspects are poorly understood of these regions and what the necessary rules are for assembling protein complexes on these regions (and often what the complexes even consist of in detail) it gets quite fuzzy.
The basics exist in bacteria and are a little more clear there due to the reduction in coding sequence. There is a paper "The evolution of DNA regulatory regions for proteo-gamma bacteria by interspecies comparisons." by Rajewsky N, Socci ND, Zapotocky M, Siggia ED that does a comparison among 5 bacteria. I think that is probably the best for promoters.
There's also a paper that discusses quite briefly the situation for intron control regions in C elegans "Conservation, regulation, synteny, and introns in a large-scale C. briggsae-C. elegans genomic alignment." by Kent WJ, Zahler AM.
There might be better papers where single genes are compared, but I am not aware of them.
DNA having a structural role is more a common conjecture that makes some sense given the way chromosomes are anchored to the nucleus, space is needed to separate protein complexes assembled on DNA, etc. I don't think anyone has directly studied it, but it is an appealing idea that is consistent with the available data.