My experience building dating apps (I built and launched a couple of my own over the years, I have never worked for a major app):

1. Men will sign up for anything. You barely need to market the app. There are at least 30 dating apps in the play store right now with at least 1 million installs and with what I would guess is around 2% or fewer female userbase. Men will sign up in droves to apps with nothing but bots and scams.

2. This means you need to design the apps in a way that attracts and retains women. You don't have a dating app without them. So men are an afterthought. This, among many different examples, is why you have height filters and not weight filters.

3. The most critical point: People say they want connections and relationships from dating apps. What they really want, shown through relentless repeated behaviour, is optionality. The dating apps that provide the most optionality, or at least the most perception of optionality, become the most popular.

With these three principles in mind, every version of a dating app simple ends up being just like the ones we have available to us now. There are lots of unique ideas about how you can implements rules and such to make an app that creates connection and relationships, (like being charged to a card on file per match, or heavily limiting concurrent matches, or only being shown a few profiles per day, or an AI that matches you) but all of those ideas violate the above principles and thus they never take off. There's a very common cope on this side of the net that it's all Match's fault and the greedy corporation is preventing you from finding love. Sure, they could do things better. Sure, they are profit motivated. But you're kidding yourself if you think an open source community maintained dating app would solve any of the major grievances people have with online dating. It's primarily a user behaviour challenge, not a software design problem.

The last thing I would say about the marriage and dating market in general is that almost every academic (economists especially) and app startup founder treats it like a sorting problem, and if only you could devise a sufficiently sophisticated algorithm, you could improve things. The truth is that it's not a sorting problem, it's a clearing problem. And there is simply no way to improve the efficiency of an unclearable barter market.