Ben,
This is something I am working on as well, I have a partial solution, but ran into a wall.

I modified an earnings date scraper written by Gus and uploaded it to Quandl via their toolbelt. The good part is that there's an earnings date time series for every stock that gets updated daily. The bad news is that it's hard to use the fetcher to get several stocks at once. The problem is that the fetcher can't grab a separate file for each stock, and the indexes don't match for each data set, so the python api throws an error if you try to concatenate them into a single call.

An idea might be to use the data sets I made to get the data for the stocks you care about, and format it into a format that can be easily grabbed by the fetcher.
The earnings date codes have this format DAVID_EDWARD/EARNINGS_{ticker symbol}

This backtest is an example with 1 stock, if the gap on MMM's earnings date is at least +-.2%, it buys or sells with 4X leverage and sells 350 minutes later. It's a start, but It would be nice to have an easy way to get this data for several stocks at once.

David

I'll let you know what I come up with. Your method is intriguing, and I may have an idea. I plan on spending a couple hours tomorrow trying to conjure up some type of solution. I feel like earnings dates could be a really useful tool as part of the API, so maybe that functionality will be implemented sometime. My other solution would be an HTML parser and just get the data myself, but I do not believe there are any supported at this point in time unfortunately. I'll report back tomorrow, and thanks for the lead.