max_buckets is useful to increase if you know that you will have a large number of elements in a given hash. data_sector_size is useful to increase if you know your data averages more than 50-60 bytes. This is how much dbm-deep allocates for every key, data value, and classname. If the value is more than this amount (minus a constant dependent on other parameters), a second data sector is created and they are chained.

To answer the next question, I have no numbers on how these values will affect performance. Frankly, I have done extremely little performance testing on dbm-deep and know of very little that's been done. I would be delighted if someone were to put together a few tests or scripts that I can have in the repository to test the effects of various performance improvements I'm working on.