When I try, I keep accidentally winning via other victory conditions. It feels like you have to play in a deliberately suboptimal way in order to win science. I'm playing on immortal, biggest map, longest game duration (I suspect that this might be causing the problem, for game balance reasons that I don't understand).
I'm currently playing this ridiculous game as Inca. I've focused almost exclusively on science from the start, campus in every city, etc. I ended up having overpowered military units, so I took out the guy who was in second place for science (was this my mistake?). This made other civs angry for some reason, so they declared wars. I could easily conquer them, so I did. Now I control an entire continent, and because of all the theatre districts and wonders that I've liberated, I'm on the cusp of a non-consensual culture victory.
I could also easily win in a few turns via domination, but I really want my first science victory. I've had to destroy all my seaside resorts, I've had to give away all of my great works, I don't repair my theatre districts if they get damaged, I've switched to a government that nobody else has, I have no foreign trade routes, and now I'm preparing to drop nukes on my own cities if they produce too much tourism (to knock out the wonders and districts).
I suspect that nuking your own cities was not the vision for how this game should be played. But I feel like, in order to win science, I have to play in ways that are clearly suboptimal. By the time I can achieve a science victory, I'm always so far ahead that it's difficult not to win some other way.
Is this user error, or is it a problem that other people have encountered as well?