You asked for feedback, and this is a great chance to give some, since I immediately saw some issues that were fundamentally hindering the simple gameplay. So while I am focusing on the negatives because they're easier to write about (and maybe because that's the kind of guy I am), I don't mean to discourage you in any way. I can see the effort is still there, and I appreciate you taking risks to liven up the game, like those 90 degree turns upwards. The neat Japanese aesthetic to tie it all together thematically is worth mentioning as well.
...
Okay, so the biggest nuisance is easy to condense into one example: those small platform jumps.
They demonstrate a lot of what needs fixing in this whole experience. I've seen straight up bugs like the player flying inside a wall. There's the classic missing the jump because I wasn't able to predict what was coming. Sometimes I manage to do it and feel like I didn't earn it.
The point about not being able to see it coming implies that the running speed in this game isn't proper for doing these kinds of precise jumps. There's no variation with how heavily you've pressed the jump button either, which is a detail I really feel you should've focused on, since - well - it's only one of two buttons, after all. Plenty of those "little things" that pile up to make me feel like it wasn't my fault I died. And they come back to haunt you on a lesser scale when fighting enemies, for example.
I think the common factor is that the game feels like it's going in fast-forward. This is something that happens with a lot of entries - during testing, the authors get bored of the slowness and ramp up some variables. After a weekend of having gotten used to the speed or difficulty or whatever other adjustable setting, they release the game, and the poor players who are just seeing it for the first time will feel left out of the ride. Having a third party playtester somewhere along the development is the best kind of reality check about what aspects of the design to focus on.
So because of the above points, you'll be resetting a lot. Regardless of how often you die, I'd say there's never a reason for a jam game to return to the title screen. The music restarting with resets is an additional small thing that becomes a big deal over time.
(In fact, I think title screens in LD games are unnecessary altogether. They don't offer any kind of meaningful representation, because a screenshot is more informative in a huge list of games anyway, and the lack of an options menu is simply an exercise in having reasonable default configurations to begin with.)
The graphics are tasteful for the most part. You've got some shader and particle type things going on that aren't overused or distracting. Be careful of pixel-perfect sprites stretching out, though. The only place where I felt like you were being lazy with the sprites was the sudden screenful of monotonous brick tiles in the tunnels.
One trick people easily overlook is fonts. Picking something beyond system fonts is an implementation-wise quick way to bring the visual style to the next level - for a game like this, many of the more readable bitmap fonts on Dafont would be appropriate. Consider the layout as well; place the relevant text to where the player can see it, and try to keep the possible text spawns to a minimum. (Text popping up in the middle of the screen is a bit of a dirty solution, and in this case makes some of it blend with the backgrounds. Better to dedicate some spot on the screen for announcements only.)
...
Hopefully you'll find at least some of my ramblings helpful. Grats for finishing and good luck on future LDs.