Time to shut down this project!
After a few weeks of actively trying to validate this idea, which included a slight direction pivot (in the same problem space), I was not able to definitely confirm the demand.
I got some verbal encouraging posts from some indie hackers, but it was just in the usual "indie hacker encouraging indie hacker" vibe.
I didn't even manage to get some verbal commitments (which are worthless, by the way) that somebody would say they would pay for something like this.
I slowly realized that the problem I thought was a huge problem was in fact just a slight annoyance that took maybe a few seconds every few days from people. Not enough to justify a 20€/month price tag. And I couldn't go lower with the pricing either due to all the OpenAI API and infrastructure costs.
So here I am now, shutting down the SpamLabs experiment, and moving on to the next idea. When it comes to ideas, I have a ton, but we all know they are worthless without execution.