Alright, gave it another shot. I think I was just pressing the jump button too early or not long enough or something because I'm able to clear stuff now easier. Though it took a few tries to realize how important the jump angle is at the very end.
Actually, it feels kind of counter-intuitive now that I think about it. You have the player pressing left and right to adjust speed. And after a few runs you realize you probably need some good speed to make it off the ramp properly so your instinct is to hold down right as much as you can and even moreso when you hit the ramp, when actually you have to pick some time kinda before or during the ramp to switch that up in favor of picking a launch angle.
The presentation is pretty rough but I can see you thought out a lot of how one screen, menu, help window goes into another so thats good. I would however remove mouse control entirely. You have the player using the arrow keys to adjust speed but you have the space bar for jump. Well that ensures that your player is using both hands really. To have to pick up your hand to move the mouse a smidge or click on its button broke the flow, especially with a game like this that seems to feed on constant retries until you "nail it". I would let the player navigate using the left/right/space config for the menus as well. Its awkward to jump back and forth.
Back to the theme then. So first, yeah it would look weird just a static baby image flying so ragdoll fits. It's not done gratuitously such as a lifeless body flung around bending and twisting limbs, etc. Ok, the point is to not throw your baby from a pram by hitting those obstacles, then you are supposed to throw your baby from the pram by hitting a final obstacle which results in the baby hitting multiple obstacles. I think there could be a market for this game but you have to choose whether its "that crazy game where you are supposed to throw your baby from a pram" or "that crazy game where you are not supposed to throw your baby from a pram". I think it may be hard to find an audience who is fine with both, even done up right. Some ideas I had to mitigate this:
-Have the Pram launch with the baby and then the baby jumps out with a parachute hitting the tower?
-A superhero or armor costume might help with letting the player know the baby is safe.
-Make the tower at the end like a pillow fort, or a blanket fort. The whole game is kind of a strange scenario. This could be a series of dreams the baby has so even transitioning the art to a bedroom/crib area at the end could help sell this or other scenarios.
-Play a sound of an adult gasping when the baby launches, but then have the baby giggle as it flies, accompanied by a sigh of relief from the adult.
You don't have to water down your concept to make it more appealing, just have to pitch it the right way. Good luck. I'm also learning a lot about construct and HTML 5 by making my first game and of course playing everyone else's here.