In fact, it's not. If you miss your renewal payment for, frex, Safari books, they actually slip your cycle date to when you renew -- since you don't *get* the service between the expire date and the renew date, I concur with his appraisal that you shouldn't be paying for it, either.
If in fact, the service *kept working* for a short time when an overlooked payment was missed, it would be a different story.
But, effectively, he's a new client, and should probably be treated that way. Assuming the paid service is actually *the update service*.
I also disagree with your proposition that this is off-topic for NANOG, really.
I've always strongly felt that this was a rather foul business practice, wherever I've seen it. The justification for it is the utterly misguided belief that, if allowed to, customers will pay for a month then cancel their subscription and 'coast' on the 'current' version of the signature for a year. This approach suffers from (at least) two fundamental flaws: 1) The entire customer base are treated as hostile. It is no surprise that they resent this. (Assumption: having resentful customers is bad) 2) Spam is, perhaps moreso than ever, a rapidly evolving threat. The effectiveness of signatures declines dramatically with time, which means that August's signatures have little value by December. [By the way, it seems to me that if they're willing to charge for valueless signatures, that represents either A) doubt as to the value of the current signatures, or B) disbelief in the decreasing value of out of date signatures.] While I realize that car insurance might not be the best analogy subject, imagine if you put your car on blocks, went off to college and allowed the insurance to lapse whilst you were there. When you return, the insurance company wants you to pay the last three years of insurance in order to reactivate your policy. That companies customers would react in the same way: they would find a new provider to do business with, rather than pay out for a valueless bit of smoke and mirrors. Nathan Eisenberg