I'm only 45 minutes in, I basically just finished the tutorial cognitive test area, I might play more later but I have some notes. So far I'd say I'm enjoying it, there's a lot of good ideas here for sure, but it does feel a bit lost in limbo between being a metroidvanian game and a linear puzzle platformer. I'll circle back to that, but I can start by talking about the segments I did play.
I really liked the hub area at the start, with the different sodas you can interact with, it's a fun space to kinda play around and experience a taste of what the game is gonna be about. Then I started on the tutorial puzzles which were... idk, fine? Pretty basic so far, nothing really cognitively engaging, but it's the tutorial so whatever. I think every puzzle involving the gravity panels is analogous to some puzzle in Portal 2 with the tractor beam things. The jump pads weren't really being used as a puzzle element in any capacity, you just use them to get to higher platforms so it's really tame compared to the launch pads in Portal 2 which interacted with all the physics objects so it was better integrated. The pink cube was kinda underutilized there, I liked the puzzle where you had to place it on a platform then move the platform to deactivate a pressure plate remotely, but the other puzzles involving the pink cube were uninspired. Overall the area was okay, I was intrigued by the potential of the mechanics but it was definitely longer and more drawn out than it needed to be.
Personally I think you're starting on the wrong foot though, loading up the game I was really excited to get my gravity flipped and have to solve puzzles by walking on walls and ceilings and stuff, so it's a bit of a drag having to go through 20 minutes of cubes and buttons. Not to mention it's spatially confusing, since a lot of these puzzles involve remotely manipulating gravity fields, but sometimes it's hard to see or tell what's going on or where the cube is going because it's in first person. The puzzles I've played felt like they would fit better in a 2d game where you can see the whole game state. Portal and some other 3d games have those puzzles sometimes where the character is locked in one spot and must remotely interact with a "puzzle wall", adopting that viewpoint would help too, but honestly I'd rather just have puzzles that take advantage of the first person view by making the challenge "how do I get there".
There's also the matter of the game feel. I dunno, I see some other commenters saying the movement feels good, but I'd have to disagree. As soon as my character jumps it's like I'm a bar of soap, you have so much air acceleration but no deceleration at all that the act of like jumping onto a box is really really difficult, I keep overshooting it or undershooting it. If you fling yourself in one direction with a jump pad, your air control is so strong you can fully reverse your momentum and slam into the wall behind you, doing an aerial U-Turn, but small corrections are impossible. I also noticed some odd interactions, like if you jump into a wall and hold towards the wall you kinda stutter fall to the ground. If you jump shortly after landing you actually jump higher, maybe this is intentional but it seemed like maybe a physics bug. I also occasionally got stuck on some geometry, like stepping onto a jump pad or pressure plate.
Lastly the metroidvania stuff... The rooms are huge, the passageways look similar so it's hard to get your bearings. I appreciate the purple line guiding me on the critical path, but once I walked off the purple line and got lot lost for like 5 minutes trying to find my way back. There were too many directions I could go, and not any obvious roadblocks like I would expect in a metroidvania game, it was completely overwhelming. Your game desperately needs a map, with like explored room tracking, show connections you haven't explored and places you haven't been. A map is really important in any metroidvania but I'd emphasize that in a game where the gravity can get all screwed up you need a really really good mapping system or players will get lost constantly, our brains are not built to navigate "true" 3d spaces like this.
Also level design wise I'd typically advise against the hub-and-spoke model for a metroidvania game, it tends to drain the fun out of exploration and figuring out where your powerups can be used if areas are sectioned off like that. Metroidvanias benefit from having less structure, or unintuitive structure that is revealed over time, or even using hub-and-spoke in smaller doses like within an area to create grounding at a local level.
And a critical missed opportunity in your metroidvania was with your first powerup. There's nothing at all wrong with a good double jump, but the reveal is kinda wasted here. In a metroidvania the delight of acquiring a new powerup is not just becoming stronger, but also imagining all the places you can use it to do new things. By positioning the double jump at the end of the tutorial area, there is none of that. I should be excited to go back and use my new powerup to reach several ledges that were out of reach earlier. Like the entrance to "area 2" should be backwards and up high, not forwards and through a ground level doorway, as well as other secrets sprinkled around in places I've already visited. Even something as simple as having a soda high up on a ledge in the break room would drive intrigue. But honestly, I didn't even expect a reward for finishing the tutorial at all, just opening the door to the main area and "unlocking" the game would have been sufficient progression.
... Anyways, don't take the criticism too hard, I tend to have more to say about games I like and that I think have a lot of potential. If time permits I'll play more of your game later... and I'll try to write more concisely next time. :P

