A solution needed:
Where there are many, and growing, downloaded images of Census returns and I am finding it difficult to settle on a consistent system for naming and filing on my computer.

Suggestion

I don't include an individual's name in the file names. I download the entire census page image, and name it accordingly. One image file can have entries for multiple families of interest, plus the file is of a census page--I decided my filenames should reflect that fact.

Example:

00-WI-Waukesha-Waukesha-ED141-i18

00->1900 US Census Population Schedule
WI->Wisconsin
Waukesha-> Waukesha County
Waukesha->town/city/locality of Waukesha
ED141-> enumeration district 141 [not used in earlier censuses]
i18->image number [taken from whatever online collection I used, in this case Ancestry]

In the 'Comments' I store the date downloaded, the online source and citation info given by the online source . For this example, that is:

Ancestry 19 Jan 2008 Year:*1900; Census*Place:*Waukesha,*Waukesha,*Wisconsin; Roll:*T623*1822; Page:*9B; Enumeration*District:*141.

The census image files are stored:
folder for all census images
-schedule folder
--state folder

So far, I haven't felt the need to create county folders. The primary criteria for grouping the files is how quickly I can find one later. People may differ on what works for them.

The individual's census citation references the specific household on the census page. The citation includes all the information I need to construct my filename (year, state, county, locality,additional locality info, image number), so I know what file holds the page image.

While doing the lookups and saving images, I save the filenames for a family/individual in a worksheet, so I know what file to use later for transcription/analysis, etc.