yes and no - in principle you find everything: regions where it plays a role, regions where it is not important, so it depends...

Hi!

In a high-throughput work by Sharon and coworkers ( http://www.nature.com/nbt/journal/v30/n6/full/nbt.2205.html ) it is crearly shown that for budding yeast transcription factors, in general, there is a "clear relationship between expression and binding-site multiplicity, as well as dependencies of expression on the distance between transcription-factor binding sites and gene starts". Plus, there are publications about specific transcription factors stating more or less the same.

Hope this will help!

Hello,

regulatory elements or motifs are part of promoters which can be spatio-temporally expressed  and I think it is also related to the chromatin structure and nucleosomes positioning which are differential for different genes. So, thinking about motifs as part of naked DNA is a mistake and we need to look at the big image

You can compare the location of these motifs with known data for histone modifications *in this cellular context* and find if they fall in "active", "repressed", or "poised" segments of the promoter - in a simplistic way, respectively enriched for H3K27ac, H3K27me3, or  H3K4me3.

In addition, you can infer from the overrepresented motif which transcription factor is likely to bind and regulate the expression of the gene. From here, you can investigate if such TF is known to be a transcription activator or repressor, its interactions with other proteins, etc.