Cutting your life line?
Okay, now about closing off your project and only distributing it to those who are willing to pay. I think this is really a bad idea especially if your project is not already widely used. The motivation for doing this is to make a premium on a per user basis. This could be done with OS software, but you have to be very careful about the way you go about it. For instance, you could not stop someone from just redistributing your work after they payed to receive it once. There is no legal ability to stop them, your project is open source and can legally be redistributed as long as the distributor complies with the license your software shares.
This may be factored into your model, and you may be willing to deal with the potential loss of revenue. Yet there are other reasons to avoid this model.
Your user base will evaporate
This is obvious. Once you stop making your applications freely available very few people will use it. And, if they buy an official release this time, they are likely to seek another distribution later. Even RedHat has to contend with this issue. Yes, RedHat does have a profitable business, but it has a barrage of clones redistributing its labor in all but name on which it keeps very close tabs. You, my friend, are not RedHat, and you will have the economic and legal recourses to keep their trademark held tight.
Your web exposure will be severely limited
Having a freely available project is a great way to get exposure. Once you close it up, expect the referrers from blogs and news outlets to dry up with your user base.
You will not sell as much as you think
Unless you have a big marketing budget, very few people will hear about your project. This of course means that you will have a much smaller audience than you once had and of course less sales.