You can spend forever polishing something  so I'd just concentrate on getting it done. Remember, you can always patch it if you find something that's really bothering you.

I think there's always an element of apprehension (or even fear) when releasing a game, and sometimes needless polishing can simply be a way of putting this off. I'd just go for it, and best of luck! 

You should aim to a demo asap, without going for perfection... it would delay the release time and so the chance to have some feedback from players, if turns out you might have to change this or that, then no amount of polishing will reveal worthy for the extra time

I basically design the fully, or almost fully in a document where from I split everything that has to be done in small independent tasks: Player controller, enemy ai, lights switch, etc. 

My basic rule is: 'it can be a little janky/buggy as long it doesn't negatively effect the gameplay/experience.'

For example: If the enemies are too fast or shoot with incredible good aim, performance issues, bad UI, etc. That something i fix. If the enemies sometimes get stuck trying to walk over decorations or their AI is so lame they just run straight to their deaths. That's something i often don't fix because the Jankiness more or less make the game more easy to the player... and sometimes is fun to see stuff like that. I mean I may fix those stuff, but only if i see it necessary. IDK, i just test things like my life depends of it and if i am happy...ish with the overall result, i just keep it like that. 

 That because I am a lone-dev and can't spent all my time and resources overpolishing everything, specially if we consider that many things that takes a lot of effort are barely noticeable in the final product, or to put in other words. Any player will actually appreciate how much time, effort or investment took you to create every single aspect of the game, they will see something that probably took you months or years and barely pay attention to it. Or maybe, arely, some will say. 'oh, that's nice' and continue like nothing. Not because they are bad people o anything, but because the only person that understand how long and how much took you to polish every single aspect of the game is the people directly involved in it. In this case, you. .... where was i? oh yeah.  Just,  go easy on you. It's okay if some things are not perfect, in the end, if 'it's just works' that's good

It’s the trap of getting stuck in the gap between vision and reality. Often the thing you have in your head when you start, and the process of making it happen are two completely different things.

When you first start a new project, it’s a perfect, shiny new idea with conceptually flawless graphics, perfect mechanics, and what you honestly believe will be an engaging story (if you have one). As you start building, we often jump feet first into the fun stuff, and then we inevitably come up against the reality. Tedious work, tuning stuff, semi-functional systems, annoying bugs, missing assets, placeholder assets, inconsistent artwork, scruffy and jagged around the edges.

This is when most devs jump ship to a new project, or segue into easy things, or polish polish polish one specific thing, because that lets us escape any sense of impending failure and get back to that cosy new game idea feeling. If you do push on, you end up looking at other people’s games and measuring your shitty first draft against polished games made by a team of people with years of experience, a reasonable budget, etc, and you start to sink again. Your own work feels insignificant or just rough and crappy.

So I think its about goals. Is your goal simply to finish a game? Or is it commercial success? The truth is that your audience is more likely to be other devs rather than players, and your reach will be minimal.

The first playable version of your game probably won’t be very good, but is it playable? Does the concept shine through? Is it fun? Interesting? You build on that. Give yourself a roadmap. A minimal set of features. You work to achieve those, and when you get close to the end of that list, you are ready to start letting people play it and feedback on it. This is what we call scope in software engineering. You design and build to agreed scope. That’s your aim.

It’s not a question of “good” vs “good enough”, but “have I achieved the goals (plural) I set myself?”, and that doesn’t need to be a one and done deal. You shouldn’t be hard on yourself if you had to cut corners. It’s a learning process, and most importantly, it should be something you enjoy doing. Once it’s out there, you’ll hopefully get people who will play it and help you improve it, but also, as you go, you will learn and revisit those cut corners.

The first game is a bitch, but it’s just a small hurdle to get passed. Good luck.

Totally get the dilemma! I think of it like this: if it is polished "enough" to get a feel for the mechanics, and whether it is fun or not, I would rather get the feedback as soon as possible, even if there are rough edges and unfinished bits, so I can improve it. 

That said, if the tinkering brings you joy, there is nothing wrong with it either depending on your aims! And there are a lot of games out there, not so easy (at least for me ;) ) to get those eyes on it!

Let me answer your question shortly: the line between  'good enough' and 'over-polished' is hidden in fulfilling your main game mechanic. One easy way to find out the main game mechanic is by figuring out if you write in your notes "My game is about ..." and just fill in the blank. Trust your game mechanic and release it as a prototype, people will test it and let you know their thoughts. Someone told me before that he's a friend that developed a simple app for 15 years! So you gotta stick with your deadline no matter how hard it might be. 😌

We actually faced this challenge as we developed our game and I was lucky to have Senna because she was the brakes that stopped us from going into the endless loop of doing last fixes. If you're not a solo developer you can ask your teammate to have the "brakes" role. ☺️