I'll close off my discussion of the NDP's leadership campaign with another lessons learned post - this one from the standpoint of the party in administering the campaign.
For the most part, the campaign seems to have thoroughly served its purpose in allowing candidates ample opportunity to showcase themselves without going into the debt that's dogged the Libs' recent leadership candidates for years. In particular, I doubt anybody would argue that the $500,000 spending cap substantially affected the candidates' ability to present themselves to actual and potential members.
There might be some room to quibble with the timing of the debates - both in not varying the debates between days of the week to reach different audiences, and in the fact that the membership cutoff hit before most of the debates took place. But the format otherwise looks to have been a success in fostering interesting discussion without becoming too combative.
And with one major exception, the convention couldn't have gone much better. The timing of the candidate showcases was called into question somewhat, but I'm not sure it was a bad idea to hold those in the afternoon to allow extra time if needed (while also leaving more time for interested viewers to record and view the showcases at their convenience). And within the showcases, the candidates found plenty of different ways to present themselves - with the immediate shift from Nathan Cullen's monologue and the entertainment-heavy presentation by Paul Dewar making for a particularly striking contrast in effective means of reaching the audience.
Of course, the one substantial concern was the delay caused by denial-of-service attacks during the leadership convention. And by all means any future votes will have to take the probability of outside interference into account - whether by providing for better triage of the type eventually carried out to separate convention voting from home voting, or by simply allowing for enough capacity to deal with attacks greater than those experienced anywhere else. But I don't see that concern as justifying a change from a model which allowed voters to participate either in advance or on the day of the convention, as the mix of votes helped to ensure members could participate however they saw fit rather than being stuck casting a preferential ballot while still wanting to be able to compare the candidates.
In sum, then, the main lesson the NDP will need to take away from its leadership campaign administration was to be more aware of outside parties seeking to throw a wrench in the works. But it speaks well to the organization behind the campaign that it took malicious outside intervention to cause a problem that will require fixing the next time the NDP elects a leader.