Pros

Allows people to build clear and simple business models around add-ons. "I sell extensions."
May grow the add-ons extension developer community by offering a financial incentive.
Provides for stable add-on development and support. Especially important for the top X add-ons: those that people closely associate with Firefox.
Apple's AppStore model is pretty seamless, as far as UX is concerned.

Cons

Disrupts end user experience - have to pay for some extensions means that there's another part to the process.
People are less likely to try out an add-on and provide feedback.
Brings money into the volunteer-driven editors and add-ons community. Unknown what the reaction will be.
Makes story complicated. Right now "everything is free of charge" is easy to tell people.
End users probably dislike this model, developers seem to like it. More likely to get feedback from developers than end users. (is this a con?)
Fewer complex add-ons are likely to be open source so people can hide IP to drive revenue. (is this a con? we already have complex add-ons shipped in binary form or with restrictive licenses)
Puts Mozilla in the middle of (and responsible for) a financial transaction between user and developer
Could make hobby developers feel like third-class citizens

When APIs change and we break extensions people paid for, what happens?
Add-ons as an incubation for Firefox features might conflict with revenue. "What happens when Firefox competes with my for-pay add-on?"
What add-ons are for-pay now and how are those handled?
How large (in some measurable unit) is the demand for this new functionality?
What are the legal ramifications of this proposed new functionality/service?
Can we go to a tipping system (ala tipjoy) instead of using a for-pay model?

This issue is bigger than just the AMO part. We can also focus on the client side, Extension Manager, code so it is capable of supporting "For Pay" mechanisms
- Installing add-ons from password protected sites