I'm having a hard time getting started. I'm struggling not just with execution, but rather, with design. Here is my current game idea of something to create...

You play a character much like Samus in abilities, on a 2D screen. No scrolling around, just one screen per level. There are 75 levels. Each level ends when you solve a puzzle. Solving puzzles involves using your gear, and worldly obstacles, to your advantage.

An example of an obstacle would be large, solid bubbles which move around but don't pop.

Is Portal like this? I haven't played the game but have heard about it. Super Metroid meets Portal on a single 2D screen per level?

I once made a 3 level Super Metroid clone. Real levels though. I wish I still had it because the clone isn't completely far off from this idea. Showing it could have avoided people calling this idea "vague".

If you're curious about Portal, check out the trailer. I'd say it gives a fair enough broad idea of the game to get inspiration from it.

There are various other ones which may be of some inspiration, e.g. Quantum Conundrum.

It's potentially interesting, it depends upon whether there's a solid enough mechanic to give a good variety of gameplay, and if it has appropriate style/humour/story/whatever to make it stick. Portal benefited greatly from the dark humour, the backstory and the goal to escape which went beyond the simple level goals.

I've never really understood why people ask this question. I don't see how answers for it should be the impetus behind a game's development. I know there's a whole market research thing that makes sense to do before producing a game but I think that's more when you're looking at investing sufficiently into a project that you want to consider the risks of going any particular direction. Even then, that sort of research goes beyond asking "would you play this?" into stuff more like "how many people have played a game like this?" and comparing things like production costs and quality vs returns.

For about a year (I think 2010) I regularly played and commented on assorted in-browser games on a couple of sites that offered them. I noticed two things. One was that what I choose to play is based mostly on a combination of impulse, visibility, and accessibility where impulse can be overridden by a curiosity instilled in me by reputation. So, yes, if your game satisfies those things sufficiently, I just might play your game. The second thing is that there is so much garbage out there that gets published to these sites. As unpleasant as it was to come across a game that was poorly designed with horrible graphics that was a complete rip-off of the popular trend of the day, it feels like 10 times worse to realize that I don't have something up there. When my attention is diverted for even 30 seconds to play such a poorly made game that has some how gotten onto what I consider a fairly reputable site, I really have no excuse for not publishing a game of my own.

What I'm trying to say is, as much as you want to be known for a well made game, it's going to be the marketing that will be deciding factor on whether your game will be played at all (and by extension if it will make money). It's a whole other beast to tackle beyond creating a game.

Honestly, it sounds to me like you're again looking for a project that you'll be passionate enough about to see to completion as other projects have hit discouraging road blocks and been abandoned. I say do your best to get something finished and, if you can, ensure that everything (or as much as possible) you do can be transferred into future unknown projects.

It all falls down to execution. Try it, see what works, what doesn't. Then hand over your prototype and we can see if we like the mood.

What I advice you to do, is go for it.. Try making the game, and the most important thing is, try it and play it first, if YOU find it not pleasurable then don't publish it, and work on it more.