It's been written that any serious photographer shoots projects rather than simply randomly snapping away at anything that catches his eye. Publishers prefer work to have a single theme and reviewers expect presented work to be cohesive.
There are valid reasons for encouraging or even insisting on a project. it certainly better tests a photographer - what can you do with a particular idea. It gives a better sense of depth - a single good image on a particular subject might just be luck or hard work rather than talent or skill, but a whole body of work on the same theme suggests the photographer is worth paying attention to.
On the other hand, there are problems with presenting your work in project form. This came to me as I was checking out some photographers to recommend. They either uniformly concentrate on a particular subject or they present dozens of shots on a particular theme, and after the first 4 or 5 images, it became a matter of 'oh, another one'.
I don't think this is fair to the work - it's quite possible that had the subsequent images been presented first, each might have been much better appreciated. Had the photographer presented more variation in images, I might have been happier to keep looking.
One could make the argument that if the images of the project are strong enough, they will withstand such perusal, but I'm not sure that's true. Perhaps it's just my own short attention span but I like a bit of variety in the images I see. I like my Weston nudes mixed with his other images.
For my own work - I know damn well I can't count on making every image on a theme outstanding - sure I can put together a body of work consisting of good work, but I can tell you that whatever the category or project, I will always have a few favourites and anticipate that viewers are likely to be the same, even if their few doesn't match mine.
I think that perhaps part of the problem is that too many images within a project look much like each other - same view point and lens, similar lighting, same subject matter. It isn't a matter of image quality - they can be very strong - but if they are too similar to other images in the project, then when they come along second or third or 17th, they sure do lose impact.
Perhaps this says something about the size of a project or of a category on your website - I lump all my B&W industrial images together, but perhaps I'm making a mistake in doing so. Not sure yet. As the images get more numerous, the problem is only likely to become worse. I don't like the idea of multiple levels of subcategories.
When I finally get round to designing a good website, I'll have to think about this.