You don't need to overengineer your training. Consistently showing up easily beats sporadic but 'efficient' sessions. Enjoying your training is the easiest way to keep showing up consistently.
If you do wonder about how to train more efficiently, then I hope this page is useful to you. None of this was obvious to me at first; I had to learn for myself through experience. The more I've applied these principles in my training, the better my results have been.
Here goes!
Train with intent.
If you do not have a specific intention while you train, you are wasting your time. It should be something fairly simple, ideally based on something you have identified as a limiting factor for yourself. For example, 'move the crosshair smoothly in the same direction that the target is moving'.
Your intention should stay constant over the course of each session. As long as you are actively holding this intention while you play -- assuming it is appropriate for your goals -- you are using your time well. The tasks you choose should make it difficult to follow through on your intention, but not to the point of making you want to quit.
If you forget about your intention, take a short break. If you are forgetting every run, play in shorter bursts. Work your way up from 15 seconds per run or less if necessary.
If you just can't concentrate at all, then end the session. Take a nap, eat, exercise, whatever you need. Come back later. If this happens regularly, your sessions need to be shorter or you have a health issue to sort out (probably sleep).
Don't punish yourself.
As long as you recognize when your actions don't match your intentions, you can learn. Look for something you can change on your next attempt and try again. There is no point in being negative about mistakes. You should take satisfaction in noticing them.
Not acting according to your intention is not a failure. It is a part of the learning process.
Unbind the 'reset' key while you're training if you want to learn to recover from mistakes. You probably do.
Distinguish between training and performance.
A good training session will involve a lot of mistakes. Don't expect highscores. Don't try to get highscores. Try to move better.
If necessary, avoid challenge mode. Stick to freeplay and turn off stats.
Focus only on what you are trying to work on in the session. Notice when your movement sucks, and try to do it better.
Identify your limiting factor.
If you can identify the weakness that is most impacting your performance, you know where you will get the most improvement for your training time.
This limiting factor should be the theme for your training blocks. It obviously should be something that is actually a limiting factor in reaching whatever goal you are working towards. It should not be something picked at random or that you've been told is 'probably' an issue, or read about in a guide without thinking about whether it is actually an issue for you. If in doubt, get a vod review.
If you've identified your weakness properly, you can trust that working on it will improve your performance later on, when it is actually time to perform.
Set appropriate difficulty.
It's hard to do this without a lot of experience, but generally you are looking for a sense of 'doable challenge'. Something that genuinely challenges you and that you can't do without concentrating, but you can do a reasonable amount of the time when you are concentrating.
Too easy and you won't learn. Too hard and most people — which means probably you, even if you don't think so — will get discouraged and give up. Find something hard enough to challenge you, but ensure you have enough success to keep you interested. This is very individual, and is one of the reasons I recommend against using premade routines in most situations. Reframing your goal for the session is a good way to find more success.
Work on one thing at a time.
For optimal training, each session should have a single theme. This can be as broad as 'tracking', or narrower depending on what you want to focus on. For example, 'tracking overhead targets', 'antimirroring closerange targets', 'avoiding tension buildup during precise tracking', etc.
You should have 'blocks' of sessions designed around a particular theme (or set of closely related themes). These blocks should be 2-4 weeks long.
You should have several different tasks in each session, which vary on your chosen theme. Obviously, the tasks you play should be appropriate to the skill you are trying to develop — but the specific details of the task matter much less than people tend to think.
Vary your training tasks.
You want to be focusing on a single theme, but you want as much variation as possible in your training tasks.
You want to minimize context dependency and maximize transfer to new situations.
You might have one specific application of the skill that you're training for (e.g. one specific benchmark scenario, or a very common situation in your 'main' game).
You still want a lot of variation in your practice because aside from helping transfer, it also helps retention -- your skill level will decay much more slowly when you inevitably take a break. The tradeoff here is that you will take a little longer to notice your progress -- but if you're training for the long term or for another game, I think it's worth it.
This post doesn't feel complete, but I'm going to publish it anyway because otherwise I probably never will. I hope it's useful to you in its current state. I'm sure I will be updating it periodically. Please reach out if you have any questions or criticisms.