Just thinking off the back of what someone else said earlier; I think the current interaction between completed legacies and age progression contributes to the problem, massively.
Specifically, it's strictly linear, and on global scale. A civilization completes cultural path? 5-10-20. Another civilization completes a cultural path? 0-0-0. Same civilization completes science path? 5-10-20. A different civilization completes it instead? Also 5-10-20. So you end up always going for culture in exploration, because, hey, someone will anyway, so it's free, might as well. I think that should be changed. For example, if all civilizations in the world only completed one legacy path, make it 5-5-10, even if they're all different paths. Once any cilization progressed two paths, the points from their second one become 10-10-20. Their third one is 10-20-40. And lastly, completing all four is 20-40-80. Put different, if every civilization in the world only completes two paths, without touching the other two at all, you get to play for 140 turns. If anyone does three fully, that shaves it down to 70. Doing all four is borderlin eimpossible without extremely heavy meta-gaming; the first two steps already take you down to 10 turns left.
That would push the games to feel different. If your goal is to build overseas colonies and heavy treasure fleets, you invest your population into rural tiles and your production into armies. Missionaries and specialists are now detrimental to your objective; if others get them, that's fine, that's their empires and their chosen paths, and because each of the empires is strong in some area, but none of them are strong in them all, the era gets to play out. As things stand, I reckon we are all now completing culture and science every time, because that's the ones AI will do anyway, and we also try to do military and economic sometimes, if we feel like it and the map allows it.