Based on... assertion? I would posit that if their teams felt they could have been comparably effective as starters over 200 IP, they'd be using them that way. A good starter is much harder to come by than a good reliever.
Perhaps the case you're making is this:
Pitcher A could be a 4.00 ERA starter or a 3.00 ERA closer. Pitcher B could be a 4.00 ERA starter or a 3.75 ERA closer. Therefore, the team uses pitcher A as a closer. It may not maximize his individual value in a bubble, but it's the best move for the team. Obviously that's simplified, but essentially the notion is that the guy would be harder to replace as a very good closer than as a pretty good starter.
There's definitely a bias in the .5 ERA adjustment as teams don't make starters out of guys who they don't think have a chance of succeeding. So you can't fairly just take a reliever and add half a run to his ERA and claim that's how he'd be as a starter.
I don't know about the other guys, but I know there were concerns about Joba's ability to get deep enough in to games a starter. There was also a concern that his stuff dropped off significantly in that role. In some ways, he reminds of Chapman. You're not looking at your standard ERA adjustment but rather a much larger one. From a process standpoint, I think you see how it plays out pretty regularly. When the guy is ready, there isn't initially a spot in the rotation for him, so he starts in the bullpen. There was always some doubt about his eventual role, so there's not really any push-back. He does well there. In subsequent seasons, rotation opportunities arise. However, the "fit" questions haven't gone away and now there's a risk associated with the move, as you have to bring in somebody to fill his spot in the pen. So unless the guy blows the doors off in a limited trial (ST, spot starts, etc.), the team sticks with what it knows works.
I'm as liable as anyone to assume I know better and that said reliever could have been a good starter. And I cite the above dynamic as evidence that the team is making a decision based on risk-aversion rather than value-maximization. But I think it's probably worth giving teams the benefit of the doubt. The Red Sox were desperate for SP, had a closer in waiting and had a very smart brain trust. If they had any belief that Papelbon could have converted and performed well, I think they would have tried.