I would say that the question is not how many bosses we think are too few or to many, but rather what is the, for lack of a better word, "flavor" of your game going to be. Both in the sense of story & in game play.
Are you wanting bosses to be largely undefined uber variants of normal monsters, or intelligent antagonists whose motives & goals are well defined?
Are you wanting to have "secret bosses" &/or "optional bosses", or are you only wanting bosses to be found along the main storyline?
Are you wanting your bosses to be what you find at the top of the proverbial mountain, that you are knowingly marching towards; or will they be challenges potentially lurking around every corner?
Most of these kinds of questions really don't have "right" or "wrong" answers, because what best serves your purposes will vary dramatically based on the kind of game you are trying to make.
For example, my current project is a sandbox game, where the player builds their team according to their own specs, with multiple endings based on player choice. Player freedom is very much a central goal of the gameplay, & player choice is a central story theme.
As a result, some bosses are conditional based upon allegiances chosen by the player, some bosses are necessary to complete the main story regardless of allegiances, some bosses are hidden because interaction with them isn't required by the player but can be hunted down if the player explores enough, & some bosses are subquest goals that are just uber variants of normal monsters.
So an uber monster hunt dungeon only has one boss, because the goal is to go into that dungeon & take out the uber variant of that monster & collect the loot. It's a diversion for the player & a reward for their exploring off the main path, so I keep it simple.
A central boss dungeon would likely have multiple bosses, because those dungeons are story focused. With the player confronting the central boss at the end, along with a conditional faction 1/2 boss, because the player sided with faction 2/1, somewhere in the middle.
As a rule the better question to ask when it comes to topics like these isn't "Which is better?", but rather "What did you do & why?". Because seeing what people were trying to accomplish, & how they go about it can be very enlightening.