Negative Space?

Figure 1 - San Francisco, California, 2011. 

Figure 2 - San Francisco, California, 2011.

Figure 3 - San Francisco, California, 2011.

Above are 3 photographs taken in San Francisco for my Urban Pieces project.  One question that I have about the compositions for this series is how much – if any - of the frame should contain, or be allowed to contain, “negative space” – i.e., not a surface painted with graffiti or a mural, and not a significant component of the urban landscape (I consider people, cars, bicycles, fences, etc., to be significant components of the urban landscape)?  I post the 3 examples above because they illustrate this question relative to the amount of sidewalk that is included in the frame.  While the sidewalk is clearly a component of the urban environment, it doesn't hold the same level of visual interest as the other components I listed above and thus I consider the sidewalk to be negative space in these compositions.  The aesthetics that I’m striving for in this series include dominance of rich colors, brightness, sharp focus, “baroque” complexity of shapes, subordination of identity of people if included, etc.  In moving from top to bottom in the images above, I ask:  am I losing anything relative to these aesthetics as the amount of sidewalk in the frame increases from 0% to about 25-33%?  I’m personally not ready to cut out any of these photographs from the series and indeed I don’t need to yet.  But this is one of the questions I would need to answer if for example I had to edit out one or two of these images.