If running a murder mystery half the fun is the players debating who's guilty and who's hiding something and why. None of what you respond answers the question: what value does it add to the game if the players can eliminate Chuck as a suspect simply because the DM doesn't ask for an insight check?
Why is that a problem? Is Chuck the only suspect? Couldn’t a roll likely also result in the players knowing he’s not guilty? chances are he’ll get eliminated as a suspect at some point, right? So what does it matter of it’s done by a roll or not? If that’s the extent of the mystery, then it’s a poorly presented mystery scenario.
Now, that’s not to say that the only way to handle it is by not rolling. But it depends on the context, and it’s up to the GM how to handle it. There are benefits to both rolling or not rolling, depending on what you want.
If the entire scenario rests on the players being unsure of Chuck, then I’d say having one binary roll that rules him out as a suspect is poor design. I’d rather have rolls that potentially reveal more information to the players without confirming or denying Chuck’s guilt. So if Chuck has some stuff to hide, but isn’t the ultimate culprit, but whatever he’s hiding may shed new light on the investigation, then that’s what I’d want to highlight, rather than just his guilt of the “main” crime.
If the entire scenario doesn’t depend on the players being unsure of Chuck, then I don't really see the significant difference between just letting them know he’s innocent versus calling for a roll with a good chance they learn exactly that.
Again, it depends on the context. Taking my Princess Bride meme as an example… Inigo invoking his father’s name in his promise to Wesley, and Wesley immediately accepting it tells us a lot about each of them, much more than a bunch more back and forth would accomplish. It also moves us along to their discussion (which again reveals more about Inigo) and on to the swordfight, which is what it’s all building to.
Sometimes a roll will help crystallize all the buildup into this one moment… how does it go??? We roll and find out and that tension that's built up gets released. Things are resolved in some way and we move on into the new situation.
Other times, calling for rolls just slows things down. It doesn’t accomplish as much as it should, or it delays other things of interest. Very often it’s done by GMs who simply think that’s what’s necessary… they’re following the steps of play and don’t realize how much that may be impacting play.
Like what if Chuck is just some ancillary character the GM had no intention of mattering to the scenario much at all. But for some reason the players have locked onto him. They say he’s suspicious and one of them says his character observes Chuck to see if there’s any cue the backs that up. The GM says okay roll Insight, and sets the DC appropriately low… but the roll fails! Now the players think this Chuck guy is involved and the GM either has to make up a bunch of crap on the fly about Chuck so they can investigate this guy, or he has to try to actively steer them back toward the stuff he has prepped. Now, even this situation may or may not be bad; some GMs will revel in the players going off on such “tangents”. But these are all the kinds of things that need to be considered.
Part of this is, I think, because the book doesn’t go into this kind of analysis. It gives very basic advice that essentially seems more about making sure no one thinks they’re “doing it wrong” than about actually talking about what decisions GMs will need to make and why. Writing to placate an audience is a bad idea, and I think the 5e DMG is full of examples why. They seem more worried about telling 45 year old DMs that it’s okay to fudge dice than they are about talking about what that can mean for play to DMs of any experience level.
This is where strong examples paired with analysis of those examples comes in handy. And that’s not something offered by Critical Role and many other videos. There’s certainly nothing wrong with watching Actual Play streams to get a sense of the game… but it’s not like Matt Mercer calls a timeout and explains his reasoning for deciding not to call for a roll. So such streams tend to be examples with no guided analysis. You see a GM in action, but without being privy to his decision making.