If an encounter says you receive a common item or curio, you don't have to pay for it unless the encounter specifically states otherwise. If you're on a space with a common item/curio icon, you can't just look at or purchase those items without resolving an encounter.

My idea: If an investigator is engaged with more than one monster, make an attack using the worst combat modifier of the monsters engaged with. The number of successes can be divided between the monsters as the player sees fit.

It's been a while since you posted and I'm not sure if you got the answers to your questions elsewhere, but here they are: 1) You're correct. If a character attacks and defeats a monster, that character takes no damage from the monster. 2) You're correct. 3) You're correct. A monster that isn't defeated by a character will damage that character during the next monster phase unless another character moves into C's space during the action phase and engages the monster by attacking it. If the monster is defeated then, it does no damage to either character. If it's not defeated, it will damage the character it is engaged with (in this case, the other character) during the next monster phase.

The difficulty really depends on the scenario. Scenarios which require the players to defeat an elite monster to win favor a lesser number of players whereas scenarios which have an "instant lose" effect if doom builds up too much favor more players. Still, even with the "instant loss" scenarios, a lesser number of players can still win if they prioritize collecting clues and performing other actions to win over reducing doom, which should be done on an "as convenient" basis.

My copy of Dead of Night came with four stands, but my copy of Dark Waters had none. I guess to save costs, FFG decided not to place plastic stands in expansions. Instead, you have to swap out investigators from expansions with the core game's plastic stands.

Here's what I do during the Mythos phase: at the start of the Mythos phase, count how many tokens are left in the bag. If there are less than two per player, shuffle all previously drawn tokens into the bag before each player draws. Remaining tokens therefor have a lesser chance of being drawn, and this also eliminates the chance of the same token being drawn twice per round. I find this technique makes the game a little more manageable for players regardless of how many are playing.