What is the ethical principle in ‘permissive’/non-free software licenses? Does it make sense as a model we should follow? No. It shows a shortfall of reasoned assessment (or even a deception perpetrated on programmers).
There is a popular rough division of software licenses into: GNU-GPL-like ‒ requiring modifications to be freed too; and ‘permissive’, like MIT and BSD ‒ allowing modifications to be kept private.
Those supporting permissive licenses appeal to the likely wider distribution such software will achieve. There is less duty asked of recipients, so more will use it. They say of GPL usage: why share if no-one looks? The point is to have the work used, so a more permissive license is better.
But this is not just about you individually: in using a license you are willing a larger pattern of many other participants. That is what should be compared.
What is the permissivist proposition? This: I should share, but others should not. In fact, worse: the poorer should share, but the richer should not. Look at the facts. You share by creating and releasing, and corporations who use your software do not. That is the result.
But what is the ethical proposition of sharing, generally? Its essence is to move abundances to scarcities ‒ that is the basic arithmetic of its ethical structure, without which it would be functionless in producing positive material effect.
So ‘permissive’ software licenses ‒ unlike GPL ‒ are exactly opposed to what sharing in principle is. They say they share, yet they enforce an anti-cooperative flow. They are self-contradictory, and so incoherent, and cannot be a rational, effective choice.