Test, by 
I think this level deserves a higher rating. It does a lot of things better than most tests available. The design is great, even if at the core it is not a perfect first time experience.
What seems difficult for any test designer is to make the difficulty fair. This test also introduces a completely new mechanic, that does not feel like regular jj2 at all. You have to also keep in mind that as a creator you get to practise it a million times while making the test and mechanics, so by the end you need to judge the difficulty through the perspective of new players learning. Assessing that situation can be difficult, sometimes impossible to please everyone.
As a newcomer to this test while being fully experienced in jj2, be ready that this is a test where your jj2 knowledge and muscle memory will not give you much, and your first completion might take several hours, while your 5th completion can be in a matter of minutes. I will not lie that my first run was quite frustrating. But this also proves a point that while it is new, it is quite consistent and not the most precise thing ever, not in comparison to jj2 movement. It just takes a lot of practise. Your first attempt failing is your lack of experience.
However, if frustration hits new player experience, does that matter? Is this a fair difficulty spike? JJ2 basic movement is very difficult to learn in itself, especially when it comes to high precision for tests. There are still tests out there for vanilla movement that are way more fit for masochists than this test, but the fact that people maybe put thousands of hours in jj2 makes taking the next steps a bit lighter. Keep in mind that for a new jj2 player 99/100 tests are completely miserable. But if we assess the test after practising, you can bet that Switcheroo is not more difficult than the most difficult vanilla tests, and this shows with how reliably I can set records in Switcheoro while still getting stuck in jj2 movement tests after thousands of hours.
So it is not that Switcheroo is inherently bad, but it will feel unnatural at first. And not only that, it feels mostly the opposite to jj2. Many jj2 controls are normally quite floaty and have delays. Switcheroo has character switches on frame 1 inputs, this clashes a bit with what JJ2 is and might feels conflicting.
So I would say what does not feel perfect about Switcheroo core design is that what it asks you to improve better each level in big leaps. But I also understand the creator wants this concept to kind of have a beginner and an endgame in one level, while also not making it too long. To have an entirely new concept in a level means that the difficulty will ramp up, and for any tests these can be frustrating roadblocks to get stuck. Which is frustrating for any tests. I think this is also a design roadblock. There is just no way to introduce such a new and opposite mechanic, fit for one level, that is both beginner friendly and poses a challenge for endgame aswell. Maybe some kind of New game+ mechanic where it unlocks more difficult levels after finishing it feels more rewarding, but that is not how tests are inherently work either. There is no solution.
Besides this core flaw, Purplejazz (the creator) did everything I think to make sure the test would be appealing regardless. There is definitely a lot of deliberate design here to ease the pain a bit. It looks visually cool, it has nice music, levels follow a clear pattern and even if you get stuck there is help tips so you know what to do. The switching is very difficult but always responsive, the new UI is very clear. During testing, Purplejazz recognized that he had a lot of practise and levels may be too hard, so basically levels were made to be easier just to fit to streamline the experience. The test does not go on endlessly with too many levels. This has less than 25 levels, so you can not stuck in a level 40 out of 50. Meaning you can still somewhat back out and not get too invested.
Personally, I also want to emphasize that whether tests are puzzles or just pure practise is a design choice, so I hope everyone recognizes this. Personally, I am more a fan of just practising levels than having to solve them first. I think this is very subjective so I was not going to rate on this anyway, but it is still worth mentioning it is a deliberate design choice.
In conclusion, I think the difficulty in this test spikes quite hard around midway through. Beating the first levels is not guarantee you can beat this whole test. You might get stuck halfway. Yes, this test is hard. Getting overwhelmed in difficult tests and having to give up is true for almost any test by nature. Yes, this can be quite discouraging, but difficulty ramp is a necessary evil to fit an entire concept into one level and to give replayability (because the last levels will remain a challenge). The creator obviously wanted to make a full experience out of this concept in a single level. Most of the challenges in this test are inherently spending a lot of time learning a complete new subset of physics. Most of the reason I am not giving this a 9 or higher is because maybe the idea itself is just a bit too much and not very fit for an accessible one level experience, but I can assure you that with these parameters set, almost nothing lacks in execution. I think the creator pushed it as far as he could. And there is obviously a great deal of deliberate and smart design decisions. Switcheroo is a level with own identity, for better or worse.
