Naturally, not having a tactic trained should not make it completely ineffective.
To put numbers to the idea: If a tactic is now 100% effective without training, the effects would be noticeable, but its effectiveness would drop to, say, 85%, after three weeks of training it would rise to 100% and after 2 more weeks to 110%. On the other hand, its effectiveness would decrease by 5% for every two weeks that you have not trained or used it, without ever dropping below 85% and without dropping anything if you use it in matches.
In this way you could have two to 4 trained and functional tactics like now or 2 highly trained ones that would work better than the current ones.
If you train for three weeks, a tactic, even if you have not used it, will be ready to use in a match with the same effectiveness as now. Let's be honest, depending on your squad there are really only two or three tactics that can be effective, if you don't have good passers the slow ones don't work and if you don't have rebounders the 2-3 doesn't work to catch more rebounds. With training they could work better and even surprise rivals more if they have never used it even if you have secretly trained it.
Advantages that I see:
I would not limit the tactics to use.
It would improve the effectiveness of tactics that do not work now.
Encourages long-term match preparation.
And above all, the game would get a little closer to the reality of sport.