You are partially right, in that its hard to get anyone to understand stuff like cryptography in the first place, let alone to apply effort in to making it better by review and bug-fixing.
But you are also wrong in two very important ways:
Firstly having something open makes it a bigger risk to back-door, and certainly makes it very hard for anyone to be offering it and not to have the come-back if they were the one putting in back-doors or spyware. With COTS stuff you have to take it on trust, which is low these days, or to try and intercept all communications with wireshark or similar and to decode/decrypt them to find out what is happening. Are people willing to do that any more common that those willing to review open source code?
Secondly the idea that open source is not needed any more is utterly wrong, as today perhaps more than ever, we are seeing a "walled garden" approach to machines where some company decides what you can do with your own hardware, and what others are allowed to offer you. Similarly having data open only works if (a) the format is published AND correct, and (b) you have access to alternative software to make use of it. Without FOSS options there would be no pressure on the likes of MS, etc, to even pretend to offer open standards and protocols.
Remember how it took an EU anti-trust suite to get the SMB/CIFS protocol opened? Or how Oracle tried to sue Google on the basis that APIs should be copyright and thus no one can make interoperable code without a license?