First off: this script is a fantastic tool. You've done an excellent job with it, to the point where we often recommend its use to new moderators.
That said, I'm becoming increasingly pessimistic toward the notion that handing folks a set of canned comments and expecting them to be used responsibly can work. At some point, it stops being a polite helping hand, and starts being enumerated badness with which to beat people over the head - or used for purposes completely contrary to the original intent. All too often, any attempt to actually help a new user understand the problems specific to their question or answer is abandoned, the canned reason becoming, in the words of ol' Pop Demand, the least they can do.
I think the best example of "canned comments" on Stack Exchange are found in the close reasons: a finite set of explanations for why certain questions must be either fixed or deleted, usable only in one context, and only with the participation of several users. A similar system can be found in the rejection reasons for suggested edits. I would like to expand that into more areas of Stack Overflow, but only when:
- There are a small, fixed number of possible comments.
- There is a specific, easy to identify context for those comments.
- The comments can be written to provide specific, constructive advice.
Again, I greatly appreciate the time and effort you've put into writing this script, and will continue to recommend it to people I know can be trusted to use it properly - but I don't think it has a place as a general-purpose feature of Stack Exchange.
Update: I'm marking this status-completed. Even though the actual implementation deviates significantly from what was requested, I think the spirit of the request is honored - a set of canned comments, for use on answers in the low-quality review queue, is now built into all Stack Exchange sites.

The exact comments offered are based on context:
- The reputation of the author
- Whether or not the author of the answer is also the author of the question
We may well expand this in the future, to other queues and other contexts.