The problem is that it makes compromises harder. It's common to say, "I am OK with your motion but it does X harm to my constituents. If you add Y to the bill I'll vote for it." That's a pretty reasonable and common way to reach a compromise, and it doesn't work if other people start splitting the question.
Adding more parties wouldn't really help. The bill still needs 50%+1 to pass. You can do that with one party, or you can assemble a coalition. But either way they're going to make compromises, and splitting a bill makes compromises harder. So US legislatures usually don't have a motion to split.
Still, the UK manages. And so do all of the organizations operating under Robert's Rules. So perhaps it would work.
Note that "earmarks" were effectively banned from US spending bills in 2011, though it looks like they may be coming back this year.
I make no opinion on the ban being a good or bad thing here, but it does look like it has some unintended side-effects.
https://www.msn.com/en-us/news/politics/earmark-ban-didnt-re...
https://www.reuters.com/article/us-usa-congress-earmarks-idU...
As to fixing the underlying problem, I've been advocating for a benevolent dictatorship by a race of alien superbeings. It neatly avoids the political turmoil and will hopefully result in cheap, easy to use jetpack-based transportation.
The issue is that we have too many people per representative.
You representative doesn't need to care about you, they need to care about donors for mass media.
I see parallels in tech as well. Some companies are openly anti consumer but run massive advertising/brainwashing campaigns. One will go as far as saying 'privacy' but turn around and give your information to the US government and China.
Representation ratio is much discussed, but unlIke FPTP there's not really support in comparative study of democracies for it being a source of adverse results (poor satisfaction, reduce dimensionality of the policy option space, etc.)
Much as the accuracy of sampling is, despite common intuition to the contrary, dependent on the absolute, not relative, size of the sample, representation ratio probably doesn't practically matter much once the absolute size of the representative body gets past a certain point (and beyond a certain absolute size, logistics make size increases adverse independent of ratio.)