It’s certainly an interesting concept. If you can pull it off it looks like it could be really fun.
Alright just beat demo
I think you may have overlooked a few things.
Why would I ever play 4 player? It appears that I can single handedly do the entire thing, so why would I want to bring in 3 other players with 3 other thought processes which are going to cause inefficiencies?
In other multiplayer games the other players give a benefit. A 4 person team deals damage up to 4 times quicker, and takes 4 times as many hits from enemies until the team is wiped. Multiple players can activate switches on other ends of the room simultaneously, jump off of each other for height, kill each other for ammo, flank enemies, etc.
That stuff isn’t in your game though. What advantage is there to a 4 person team in your game?
While some people will probably call me elitist, in my time spent gaming, I’ve found consistently that having to rely on others is not fun as the odds of them dragging me down are substantial. If I’m going to take the inefficiency and risk of having a 4 person team controlling a single avatar, there needs to be some reward to balance that out. There doesn’t appear to be one.
It might be a good idea to have a 4 person team be able to split in to 4 distinct entities (See Lazer Team trailer). So one player would be a phantom with an armored arm, while another is a phantom with the boots who can dash and ram, etc. So give the team the option to split and reform. That would allow the players do be able to do things like hitting multiple switches at once, flanking, etc.
Perhaps enemies drop some currency which is then used to buy gear. If you have 1 person then you get 1x currency. If you have 2 people you get 2.5x currency split between the two of them, meaning each person gets 1.25x more than they normally would. Then for 3 people, 4.5x currency. And for 4 people either 7x or 8x currency. Then players would party up because grinding would go quicker. Though that is only a temporary solution as once the player has purchased everything they’d go right back to single player.
Perhaps each piece gets stronger as more players are there? For example, the sword could hit harder, faster, or have an improved range. The headpiece could cause damage as well as stun or stun for a longer duration. You get the idea.
Perhaps each player has their own field of influence? So the sword player only attacks right in front of the player, the foot player can move anywhere, the arrow player can fire in a 360 arc, and the head player can look independently of which direction the body is moving. Then for the single player experience, leave it just the way that it currently is so that a single player can only attack in front of them. Then you’d really see a difference with 4 people playing.
Also you should really consider each person’s experience. The person who gets the head, all they’re doing is spamming the paralyze attack repeatedly. There’s no agency, no strategy, no becoming better with experience, just mindless repeated button mashing. The other 3 aren’t as bad, but really the head player sticks out as clearly having the worst role of the 4. I get making the boss immune to paralyze, because that would make it way too easy, but you’ve essentially just designed the head player to be 100% useless during a boss fight. There is no possible way for them to contribute if you make the boss immune to the player’s only move.
What is a team going to do if they get a troll on their team? Somebody who refuses to play their role, and is only there taking up a slot for the lulz? Will you have a vote to kick system? What types of protection or incentives do you plan to have in place to minimize that? If a troll gets the ability to repeatedly drop bombs (Which damage the player), it’ll quickly ruin the experience for the other 3 players.
So right now you have a lot of the negative aspects of multiplayer, but you don’t have any of the positive aspects. Really work on reversing that. You want a lot of the best parts of multiplayer, but done in a way to minimize the worst parts.
Consider revamping the dead ends. Fighting a room full of enemies is fine if it’s on the way from point A to B. When I go to a room at the edge of the map (To explore and see if I can find some secrets or loot) and all I’m given is another room of enemies to fight, then what was the point? Just to waste my time?
The field of vision “Puzzles” were interesting, but they were enough of a pain having to slowly crawl through to do it myself. I shudder to imagine having multiple people trying to pull that off.
Lastly, if there isn’t any benefit to fighting the monsters, either in the form of currency or loot drops, then it might be easier on the player to only have the enemies respawn whenever the player revives at a checkpoint, rather than every time they enter the room. Alternatively, it may just be better to have mobs drop currency and mob specific loot when they die, rather than just the occasional health replenish. That way the rooms with mobs aren’t just filler / padding and the player can get a sense of character progression.
For single player the difficulty was a bit on the easy side, but that might wind up being good with the inefficiencies caused by 4 players.
Perhaps this just isn’t in the demo, but don’t forget that you can use the bombs for puzzles as well, and of course cracked walls to blow up are a tradition in that type of game. This also goes for the arrows and paralyze. Try to come up with a puzzle which all 4 players can participate in. Much easier said than done I know, but it would give a nice sense of accomplishment. Look at what Fable 3 does with their flit switches. 3 different colors, and the color lets you know if you need a melee, ranged, or magic attack to activate it. You could use that for sword, arrow, and paralyze vision.
I’m not sure if the concept is entirely original, but it’s still different from 99% of what’s out there. I feel that you’ve got your work cut out for you on making it work, but if you pull it off I think people will have a blast with it.
As far as “Does this game have potential”, absolutely. Pleasing art style, good AI, and it controls pretty well.
The main issue is that the multiplayer just feels tacked on as an afterthought, rather than having the game built from day 1 as a multiplayer game. The puzzles don’t require any meaningful team interaction, and neither did the combat.
Really think about what you’re going to do to convince players to play with a team rather than going solo.
Really think about how you’re going to design your game to showcase and highlight the multiplayer as the main mechanic.
I think how you address those two points will either make or break your game.
Good luck to you 
