Yes and no. I've imported 700,000 records from Excel spreadsheets, but I found that a sensible maximum number of rows to import at one time was around the 10,000 mark, so I broke down that import into 10K blocks. Also, don't forget that Excel can only handle 65K rows.

When importing from an Access database (same spreadsheet tables, only imported into Access) using the ADO connection option, then I've found I can import in excess of 25,000 at a time. I'm sure it could handle more, but the problem lies in the matching stages. If you have too many species to match, too many locations etc. then the stability of the import wizard becomes a factor. You could spend all day matching species and locations only to have the import wizard crash, losing you the matching work you've done.

So the bottom line is, do large imports using Access if you can. Mike Weideli has also mentioned to me in the past that building a separate utility for importing large quantities of data would be relatively simple. Such a thing wouldn't be as flexible as the import wizard though.