In the most recent issue of the Genealogical Record of Stafford County, N.H. [May/June 2010], someone found two issues of The Strafford Inquirer from April 22 and June 10, 1828. So the journal reprinted all the marriage and death announcements contained in the two issues. Coincidentally, I was discussing with a genealogical colleague about the coverage of newspaper databases now. Have they reached critical mass for information? How many newspapers were still out there in depositories not yet digitized? Specifically, is it worth it to travel to Worcester and see the collections at the American Antiquarian Society? So, this seemed like a good example to explore. How many of these marriage and deaths were already available?
The April 22nd issue had 6 marriages and 10 death announcements. The June 10th issue had 16 marriages and 11 death announcements. The first place I checked was William E. Wentworth's Vital Records 1790-1829 from Dover, New Hampshire's First Newspaper (Camden, Me.: Picton Press, 1995). Since the newspaper in question was also published in Dover, I expected quite a bit of overlap. However, the New Hampshire Republican (the name of the newspaper at that time), didn't have good coverage during April to June 1828. Only four of the events appeared in this work.
I then turned to genealogybank.com. By searching just New Hampshire newspapers, I found 9 of the 16 events of April 22nd and 23 of the 27 events of June 10th. When I broadened the search to all newspapers, I found 11 of 16 for April 22nd and 26 of 27 events for June 10th. Overall that's 37 out of 43 or 86%. That's pretty good coverage. Of course, if your genealogy depended on the missing 6 records, then, it's a different story. Only two missed events happened locally at Barnstead and Atkinson, N.H. The others were all from other places such as Boston, St. Thomas, and New York (2 events).
So, my conclusion is that for early New England** there is a decent chance that marriages and deaths have already been picked up, but there is always a chance that a certain paper has the only mention of the event. So in true genealogical fashion, your search is never over.
**I use these databases for the midwest for Michael's ancestry and thus far, I've found very, very little. I'm really hoping at some point the papers of Leavenworth Co., Kansas expand. They seem to be from the 1850s to 1870 and then from the 1890s on. Everything I need is between those dates!