To be fair, GoG has relatively limited resources to work with. It has around 10% of the market share of digital sales compared to Steam's 60%.
Let's say there is one guy, Craig, who does the curration. Ptarth says, 'Hey Craig, check this new thing out." Craig says, "Okay, I'll put it into my queue." Then 3 weeks later Craig is able to get through his queue and takes a look at Ptarth's entry(3 weeks? Yep, seems that way:
https://www.reddit.com/r/gamedev/comments/46o3fj/a_little_story_about_gogcom_and_my_experience). At that point he thinks it doesn't fit with the image they are trying to project and emails Ptarth telling him, "I don't think it works for us." Then Ptarth says, "But wait, check out this new version..." Then Craig has to decide if he is going to look at it now, put it on his queue, or never look at it again. It sounds simple with just 1 game, but if your current workload let's you make a decision every 3 weeks, then if you have to give a second opinion to everyone, you are now needing an extra 52 weeks per year to handle that load.
But wait, it gets worse. Let's say that the game is rejected a second time, and Ptarth wants to a third review, you now have tripled your workload. And worse yet, you've now trained Ptarth that it's okay to make multiple submissions, since he is now following a "Well, it will eventually it get through, I just have to keep submitting" strategy. In order to keep things reasonable, you do need an absolute criterion. Sometimes you miss out, but you have to consider the cost of each additional operation relative to the potential incomes it gives.
The second problem is flavor. GoG has a idea about the kind of products it wants to sell, let's say waffles. Who are we to demand that they sell pancakes instead? Furthermore, if they don't want to sell pancakes and we then threaten to bad mouth them all over the internet, that's not very fair. If they want to sell waffles, then celebrate their desire to sell waffles, even if what we want are pancakes.