Why are there 2 steps in linking a character to a token?

The way I create a character (eg. Barry the Barbarian) in a campaign is setting up the character sheet first. Next I drop the token artwork on the map and I select "Use selected token" in the Bio/Info section for Barry. When I later drop Barry the Barbarian on the map the token appears just fine, but I can't bring up Barry's character sheet by shift+double clicking the token. It's like the platform knows which token to use for Barry, but doesn't connect the dots to connect the character sheet info to that token. In order for this to work I still have to open up the token settings and select Barry in the "Represents Character" dropdown menu. So in summary I have to connect the token to the character, and the character to the token... This seems necessary to me. When I "Use selected token" in the character setup this should be enough, shouldn't it? I don't see any reason why I should have to set up the "Represents Character" section of the token when I've already connected that token in the Bio/Info section. Is there any reason for this that I'm missing? Or have I been doing this wrong for the last couple of months? It's not like it takes up so much time but I always found it weird and when I prepare a campaign and want to add a load of PCs, NPCs and monsters it can be a hassle.

Having a character's token link to the character is the most common use case, but there are edge cases where that isn't the behavior you want. Here's a few off the top of my head: This is an alternate form character, where you want a different graphic (size/shape, image, aura, tint, etc) for another character, but want all the stats and abilities of that other character in one place. Examples: Knight on a horse, Druid as a whale, mecha in jet plane mode, dragon in human guise.  This is an alternate configuration character, where you want different bars linked and auras set up, a different name, but the same image and linked to the same character. Examples: decker/hacker in cyberspace, mage in a no-magic field, space ship in an atmosphere, This is an alternate owner character, where you want to give a player the ability to place a token they don't control. Examples: player gets to pick the position of their hireling, but doesn't control them, player places their opponents pieces in setting up a board game, mage places a spell template, but doesn't get to adjust it afterwards. This is a convenience character, where the graphics isn't intended to represent a character at all, just part of a palette of peaceable objects. Examples: different light sources (torch's, bonfire, flaming floor tile), furniture, buildings, traps, map markers, corpses These might not apply to every game and there is probably room for improvement to the interface, but I'd hate to lose the flexibility. 

When you drop the token on the map for the first time, do you set it up fully for Barry the Barbarian,  Set the 'represents character' set the bars for health etc, set any dynamic lighting options up, and then click 'use selected token' on the character sheet.  That should mean that every time you drag Barry from the journal the token is already set up and ready to go, it shouldn't need to be linked a second (or third, or forth) time.

@ The Aaron I guess I haven't encountered any situation where I would find any use for it. I undestand they try and implement the features so they are useful for the majority of people. I guess I mostly wanted to check whether I was doing something wrong.... @ Oenanthe I understand what you're saying but in you're example you're also making the connection twice, no? My thought would be that if I select Barry as "Represents character" I shouldn't have to do the "use selected token"-part again, as I've already made clear that this token will be used for Barry.

If it automatically linked by using the Represents drop down, then you couldn't have 2 tokens representing the same character while being linked to different attributes of that character.  An example might be using the default token for your sheet (Bob) as the on-map avatar of the character that tracks health, armor rating, and armor "health".  Then you have another character (Bob's Casting Stats) where the default token also represents Bob, but tracks spell points, a variable spellcasting modifier, and the damage absorption remaining from one of Bob's abilities.  So now you have 2 tokens that both represent Bob, but are tracking 6 different things.  They are each set as the default token for 2 different sheets meaning that they are easy to drag out on new maps by player or GM.  The secondary wouldn't necessarily be needed all the time and would likely be set off along a map border when in use. I don't know what game system you play.  This may not be something you would ever need for whatever you are playing, but other game systems certainly benefit from being able to do things like this.  I run a system where a (fairly complex) character could potentially be tracking health, armor rating, armor damage capacity (armor health), remaining casts per day, and inner strength points that fuel psionic abilities.  They may also need to track the remaining damage capacity of some magically generated shield/armor.  Yes, those are all (except the temporary magical thing) on the sheet, but it can be faster and easier for some players to just alter the number in a token bar, especially if they are accessing Roll20 from a computer that can barely handle it. Got a game system that does mech/vehicle activity, but the pilot can do things separately as well?  If the sheet is set up for holding the stats for both, you would have a token for the pilot and a token for the mech/vehicle that both represent the character, but one will be the default token for a different sheet. The use is there.  It just may not apply to the particular game system you play or the sheet you are using.