I would say that the difference between a privacy encryption app like TrueCrypt and the examples you've used are that the ones you listed especially with the DieBold machines which centers around voting rather than privacy all revolve around the flaw of being translucent rather than transparent in their execution of their jobs where as a privacy app need not be that way. It simply needs to be 1. effective and 2. not be as you pointed out "shitware"
In fact such examples though different in category highlight what happens when people become apathetic towards the value by which they claim to cherish and then only find out later on that they have been screwed yet continued to bitch about it rather than make a difference because the technology forced them to remain clueless yet paranoid. A common result when people are led to believe in technology blindly (sometimes not even led, just apathetic to any cons technology brings which isn't help by ones own subconscious apathy towards the voting system)
I would also say that the flaw of saying something is the best at what it does must still assume that the user wants to have the best encryption program rather than the best way to keep their privacy. This leads to two major flaws IMO:
a) By saying that you are merely pointing out to what you consider the best choice totally throws out your earlier argument of Open Source vs. Closed Source:
Imho this is wrong. The only thing closed-source gains you is obscurity - and everybody who's into security is going to say that security through obscurity never works. For stuff like encryption, having the source code open inspires more trust than depending on bugs (and backdoors?) not being discovered.
Why? Because you are inherently comparing the best Open Source model in TrueCrypt and justifying the model of Open Source through that one app rather than addressing the actual model of Open Source.
I don't think that opinion is wrong and I did hint of it's popularity in my reply but it still does throw out the open source model and rather makes a case that the popular optimum model is the best choice when choosing a privacy app and that being Open Source only has relevance to it because the current popular program happens to be one.
b) By assuming it is the best choice, such attitudes (especially when it becomes one adapted by a group large enough) becomes the very influencing factor in convincing people (especially people ignorant of the backbones of privacy applications) to assume that it is the sole Holy Grail of privacy apps and yet even as you point to it being what you perceived as the best choice, you also allude to it only to being a "decent" encryption program (something I disagree with btw, I think most apps have reached that stage of being great encryption programs for their purposes) and also that there is more to security than it alone.
Words like those in my opinion only prove that TrueCrypt isn't exactly secure but only considered secure enough by the majority of tech users. Words like those also hide the fact that TrueCrypt still requires improving so even though it is the best choice considered by many currently, it is only if you enforce that belief from the software design and effectiveness perspective rather than the privacy perspective that it becomes "secure enough".