No, a sandbox provides an actual layer of security. You could argue about whether or not this or that sandbox implementation succeeds, but there is no getting around the fact that it is by definition an additional system-level layer of security.
Given the verbiage that you used, what you're clearly trying to express is, "anything that's not open source is inherently insecure." If that's the point you're trying to make, you should just make that point, don't dance around it by making it sound like you're talking about the usefulness of sandboxing.
Hell, chroot could be considered a form of sandboxing and it's used in open source all the time, as are VMs, as are jails, as are having dedicated system accounts for different daemons, as are a gazillion other technologies and practices that provide additional separation between software and the system it's running on. Surely you don't think that stuff is all smoke and mirrors to make people feel safe.
Given the verbiage that you used, what you're clearly trying to express is, "anything that's not open source is inherently insecure." If that's the point you're trying to make, you should just make that point, don't dance around it by making it sound like you're talking about the usefulness of sandboxing.
Hell, chroot could be considered a form of sandboxing and it's used in open source all the time, as are VMs, as are jails, as are having dedicated system accounts for different daemons, as are a gazillion other technologies and practices that provide additional separation between software and the system it's running on. Surely you don't think that stuff is all smoke and mirrors to make people feel safe.