I respectfully disagree.
An error was identified, communicated, and acted upon as a result of open data. Specifically, it became known that due to inadequate training practices patrol officers were unawares of the change whereas parking enforcement were the only ones trained and therefore were not making the same mistake:
> "When the rule changed in 2009 to allow for certain pedestrian ramps to be blocked by parked vehicles, the department focused training on traffic agents, who write the majority of summonses."
> "Yet, the majority of summonses written for this code violation were written by police officers. As a result, the department sent a training message to all officers clarifying the rule change and has communicated to commanders of precincts with the highest number of summonses, informing them of the issues within their command."
So the issue isn't that police were confused so much as their training wasn't updated because of a system failure (training only traffic agents and not patrol officers).
Further they are taking action to monitor the open data themselves to detect future anomalies themselves:
> "Thanks to this analysis and the availability of this open data, the department is also taking steps to digitally monitor these types of summonses to ensure that they are being issued correctly."
This is a great step forward regardless of whether or not en-mass refund occur.
To me, a refund would be icing on the cake. And perhaps one will happen (hopefully) this post was published today (May 11, 2016) after all.
As an aside, there is also something to be said about the fact citizens should be aware of their own rights and laws. Any citizen could and should have contested wrongly issued tickets.