AFAIK Steam only has "Recommend" or "Don't recommend", so the super enthusiastic simpletons' votes carry just as much/little weight as a carefully examined vote. This mostly gets around the "0/10" problem.
It kinda was though, so your argument is a bit moot. It may have been an incredibly shallow game, but it was fun and really popular... which makes it a good game!
Ratings and comments are both important metrics though. You just need to be smart about how you use them. A rating will tell you how well your game is being received, and by extension how much tweaking you need to do. If people already like your game, DON'T make major changes to it, even if people in the comments are asking for it. If you change things too much you can potentially upset the silent majority, so don't fix what ain't broke. Comments give you insight into what could be improved but you need to consider every comment with respect to the game you're making. Tons of people might be asking for a specific feature, but if you know that including it would weaken or break the game you shouldn't include it. For example, say you're making a Zelda-like and people want craftable weapons, you need to examine the impact craftable weapons will have on the game's balance before just chucking it in. People are stupid and don't know what they want.
A great example of this is Pokémon. People at Nintendo, GameFreak and The Pokémon Company have gone on record saying that they completely ignore what people say they "want" because people don't know what they want. They want all the regions, but then they only want the first 150 Pokémon to be available (because, "fuck Mew", apparently?), but they don't want ANY of the first 150 to be available because they're "tired of Nintendo catering to genwunners", and they don't want Mega Evolutions, but they want Mega Evolutions for Pokémon who don't have them and more regional forms, but not the Alolan forms because those sucked, and... Etc. etc.
Edited by Binx, 21 June 2018 - 08:33 PM.