Hi Dennis, Arbor,
In our research we have shown that crawling severely underestimates the network size. We have also — in depth — studied Zeus P2P, showing that it spans more than 100k bots. You may want to reconsider the way how you phrase your result interpretation.
Detailed results see: http://christian-rossow.de/publications/p2pwned-ieee2013.pdf
I also do not consider it a good idea to crawl the botnet without knowing the exact implication this has in terms of topology changes.
Just my two cents.