I think that your results are exactly what they should be. The committed version of the notebook was executed on Jan 22, and the default parameters model changed quite significantly since then. If you revert back to the Jan 22 revision 79b4fb5d48a5feb91eef15e02bc356d093160a59 (and revision 37a5e17d4b62a495e46fb21af0d55b07d7bf5e15 of nengo-bio) you will get the results that were stored in the notebook.
Edit: As the name of the notebook suggests, we only used this notebook for playing around with the model. The data shown there are not what we reported in the paper. We used the notebook Run Model.ipynb for that, which I apparently didn’t commit until Apr 2. It thus wasn’t included in the repository that I uploaded to GitHub. I have now added Run Model.ipynb.