Sorry for the delay in responding.
So... First thing and I know it doesn't directly help with the local effort of adding product to the spreadsheet, but EP4 offers the ability to split a file that is either exported to the server or uploaded to the server. So you could (later) do updates of existing data by exporting, then splitting then downloading so that when you open the file you aren't "troubled" by too many rows/too much data.
That said, if you haven't downloaded the github version which supports product export without including linked product, then when a set of data is split, there is the possibility, however remote, that a row of data for a product could be in one split file and another... the last one imported wins...
To your questions about uploading new product. The concept of the software is that when working with the main product information, only required fields need to be in the file. For new product that is a primary key and a category. For any field not included, generally speaking it will default to some form of emptiness. For dates, if the date_added field is not included, then it will default to now, whenever now is. If the field is included and it is blank or has a date less than 0001-01-01 00:00:00, then it will be set to that (which should be the default in the database, but that's a different story.) v_date_avail does not default to a now condition, but instead a "blank" condition.
As to product status, if it is not provided then it should default to either off (disabled) or the status default in your database. There is a consideration though as to if a product is uploaded that has 0 quantity and the store is set to deactivate product that have 0 quantity then the product will be deactivated regardless of the condition of the uploaded status. Sort of an EP4 working with ZC condition.
Otherwise, regarding your "plan" for updating. Again, when discussing the main product information (not yet incorporated in attributes and a couple other off-shoots), if you don't want to even possibly change a field, then the field column should and can be omitted. If the field is present, whatever is in that column will be attempted to be pushed to the database. If that means the column is blank, then the data will be removed, if it changed from "up" to "down" then "down" will be what remains...
Now, if I have incorrectly stated these things, such that the software does not operate as I have described, please identify as it more than likely is a mistake.
Now, attributes and category specific changes, I know attributes haven't been further modified to permit import with fields missing, but I think at the moment that category import is the same (operation hasn't been modified to support upload with missing fields).
Again, to your last question, there has not been a setting established to leave a field alone if the field is present but the row data is blank. I don't seem to recall anyone asking for that and I tend to look at it as extra work to blank a field to prevent an update as compared to possibly two different file imports or to leave the data as is. (it's possible in most spreadsheet programs to "lock" a field to prevent editing, so would recommend that route instead personally...)
