The purpose of this statistics gathering is twofold:
You are encouraged to participate in the latter. This is being done in conjunction with the Internet Engineering Task Force's DKIM Working Group, the body that created the DKIM standard. We will submit the names of any company or person that contributes data as part of our report to that body for appropriate credit.
You are not required to participate. In fact, it is necessary to provide a specific build-time option just to get the statistics stuff compiled in (and the filter works just fine without doing so), and then you'd need to take specific additional actions to submit the data back to The OpenDKIM Project. The data submitted will identify you as the submitter, but you can choose to submit only hashes of the author domains of the messages you are reporting; this way we can't tell what the author domain was, but we can tell if two different messages about which you submit data are from the same author domain. This is important for accurate data collection.
If you are interested in experimenting with the statistics features or learning more, download v2.2.0 or later and look at the README file in the stats directory.
Please submit any questions about this effort to the opendkim-users mailing list.
The latest auto-generated report based on submitted data can be found here.