No hard date. This is high priority for us. We dont have a good enough time tracking infrastructure in place to give good estimates yet. After the outrage at our inaccurate 4.0 release date estimates we wont give time estimates until we have that time tracking infrastructure in-place.
Shorter answer: ASAP.
The good news is that the client-side fraud detection in 4.0 is good right now (not perfect) and it *is* susceptible to cooked up experiments where the local clock is rolled forward when a customer is already in a trial will add time to the trial. Real world usage of TurboActivate 4.0 is better at detecting client-side fraud (because real world Time is elapsing and TA is able to collect real world data points on the actual time on the computer).
TA 4.1 solves the cooked up experiments from showing long trial lengths. Plus it also solves more edge cases of real-world client-side fraud.
Also, client-side fraud detection will *always* be imperfect no matter what product you use. Its just the nature of doing verification of a trial without *requiring* an active internet connection while they use your app. The licensing (us, in this case) needs to collect real world data points and ignore fraudulent ones all without having access to a trusted source in the real world.