I don't think it's a matter of growing out of it. Rather when I was younger I tended to uncritically accept the rhetoric surrounding the GPL vs BSD licenses. As I have watched the way many BSD-licensed projects work on the ground, that's changed.
Two major realizations have occurred to me:
1) the GPL generally provides a shield to those projects which have a slow pace of development or are immature. They prevent things like Informix or Solaris from developing from PostgreSQL or BSD.
2) On the other hand, as projects mature, there is a real quid quo pro regarding future contributions which doesn't exist in the GPL. By contributing back you ensure you have access to future patches by other contributors. As a result companies like EnterpriseDB or Green Plum which offer proprietary versions of PostgreSQL have both contributed a lot of their code back. They don't contribute everything back but they contribute everything back the project wants to accept. This leaves EDB with some Oracle compatibility code and GreenPlum with a node coordinator and a few other things.
The quid quo pro is relatively simple. if the pace of development and quality of code is high, then you really want cost-effective access to new work. The more you keep to yourself the more it costs you to access that work, so you give things back so they become standard, and reduce your own efforts to integrate what is left.
I typed a much longer response, but I realized that the internal contradiction in this argument can be summed up in many fewer words: "...there is a real quid quo pro regarding future contributions which doesn't exist in the GPL. By contributing back you ensure you have access to future patches by other contributors..." <- that is, in fact, a description of the behavior the GPL requires... claiming that it doesn't exist with the GPL thereby is confusing at best. The reason people attempt to avoid the GPL is specifically because they either currently intend to not play "quid pro quo" or because they want to reserve the right to do so in the future.
Two major realizations have occurred to me:
1) the GPL generally provides a shield to those projects which have a slow pace of development or are immature. They prevent things like Informix or Solaris from developing from PostgreSQL or BSD.
2) On the other hand, as projects mature, there is a real quid quo pro regarding future contributions which doesn't exist in the GPL. By contributing back you ensure you have access to future patches by other contributors. As a result companies like EnterpriseDB or Green Plum which offer proprietary versions of PostgreSQL have both contributed a lot of their code back. They don't contribute everything back but they contribute everything back the project wants to accept. This leaves EDB with some Oracle compatibility code and GreenPlum with a node coordinator and a few other things.
The quid quo pro is relatively simple. if the pace of development and quality of code is high, then you really want cost-effective access to new work. The more you keep to yourself the more it costs you to access that work, so you give things back so they become standard, and reduce your own efforts to integrate what is left.