In the early days of free software, the struggle was just to get companies to try this new and rather unconventional approach, without worrying too much about how that happened. That typically meant programs entering by the backdoor, surreptitiously installed by in-house engineers who understood the virtues of the stuff -- and that it was easier to ask for forgiveness after the event than for permission before.
Later, when free software become more widespread in the enterprise, the issue of license compliance gradually reared its head, particularly in terms of making code available and contributing back changes. At this time, it was still important to minimize barriers to using free software, rather than to police it in such a way as to frighten off potential supporters. But it was also critically important to confirm that the "copyleft" approach worked in legal terms and would be accepted by the courts.
This led to a very low-key, discreet approach, whereby defenders of the Gnu GPL -- because this was the main license involved -- tried to persuade companies to comply, without resorting to heavy-handed legal methods. That explains why there have been so few cases argued in court, and why those that have moved in that direction have all ended successfully for free software, as in the latest example.
As a result, it is now pretty much established and accepted that the Gnu GPL license is sound, and that there is very little point in trying to wriggle out of complying with it. That being the case, the emphasis is now shifting from slapping down the occasional willful noncompliance to making it as easy as possible to check for compliance, as evidenced by this major new initiative from the Linux Foundation: