The Iroquois are the worst because the longhouse loses out on the production percentage bonus of the workshop, while supposedly replacing it with the production from forests. The problem here is that throughout your empire, in a normal game, you would rarely work more than something like 3-5 forests in any one city, and that's generously assuming cities will have lots of deer or non-fresh water forest tiles. However by the time the workshop rolls around (usually after universities) most cities would be taking in somewhere from 20-30 production (more in the capital) meaning the longhouse would usually only be slightly better at best than a workshop right when they are available. Then when you factor in the long-term implications, such as late game cities often pulling in more than 50-60 production with the capital at least around 80-100+, it usually becomes the case that a longhouse is actually much worse than a workshop. Now that's how a normal game would go. The idea then is that the Iroquois would leave forests untouched through the empire, something the UA is meant to encourage as well, but this means sacrificing growth from farms and rush production from early chops, two crippling requirements to make the Iroquois "bonuses" work.
This means the longhouse is WORSE than a workshop, while the UA (which, as Filthy mentions, is basically broken in vanilla) also requires reliance on allowing terrible forest yields to exist through the whole game instead of replacing them with farms or other improvements. That leaves the UU, which, like all swordsman replacements, is just about useless as the swordsman rarely sees play (during this time of the game the pikeman is on a more convenient tech path).
