fetching ppa stats with a ppa.getPublishedBinaries() loop is too slow

When I get the stats for the chromium PPAs with this code, it's way too slow..

        for bin in ppa.getPublishedBinaries():
            d = bin.getDailyDownloadTotals()

Polled 89445 rows for chromium-daily/ppa (20361, 22% empty) in 3:57:31.858862
Polled 10545 rows for chromium-daily/dev (3722, 35% empty) in 0:17:39.911383
Polled 9233 rows for chromium-daily/beta (1399, 15% empty) in 0:14:39.380236
Polled 4543 rows for chromium-daily/stable (1082, 23% empty) in 0:07:12.078046

So basically, it takes 4h30m to get the stats for the 4 PPAs, and that's increasing.

Somehow skipping those empty rows would spare 24% of the requests.

Also, to do an incremental update of the stats, it would be nice to be able to pass a date filter to getPublishedBinaries(). Today, the only filter is a package name and there's no API to get the binary names beforehand (allowing me to fake my own date filter).
The current scheme reports the binary packages one by one, but iterates over all dates before moving to the next package.