You were basically claiming that we could use swans to teach students the mechanics of jet flight. I generally don't provide evidence in response to posts such as those, since the distinction should be obvious to anyone who gives the topic even half a glance.
But backing up for a sec and giving you the benefit of the doubt (since you responded more politely than my comment warranted), whether you realize it or not, your original assertion effectively says that there isn't a distinction between natural languages (of which spoken languages* are a subset) and formal languages (of which programming languages are a subset), which is wholly incorrect. While a number of concepts carry over between the two (e.g. alphabets, grammars, etc.), there are a number of other features that remain unique to each, meaning that you'll quickly get into the weeds if you try to use the one to explain the other, just the same as if you were trying to use a swan to explain jet flight.
Just consider what makes a formal language "formal": that it's mathematically defined. They consist solely of denotation. Everything is literal and can be taken at face value. In contrast, natural languages frequently make use of both denotation and connotation, the latter of which cannot, by definition, exist in a formal language.
Likewise, you can't explain the closure properties resulting from the various mathematical operations on formal languages without first delving into an explanation of the mathematics that make formal languages what they are. Natural languages have no analogous concept. Even if we restrict ourselves to the practical matters of programming languages, there are a number of everyday concepts (e.g. recursion) that don't exist in everyday natural language (cue people linking WINE and GNU).
Perhaps more central to your original comment, I disagree with the apparent implication that they can be taught similarly. After all, even if we set aside the technical distinctions, the types of thinking required by the types of languages necessitate that they be taught differently. Natural languages are very fault tolerant and provide us with a number of "recovery" mechanisms to infer meaning in cases where there's the potential for ambiguity, thus splitting the burden of successful communication between the speaker and the listener. In contrast, formal languages put the onus solely on the speaker to successfully communicate. And they either do it, or they don't. 1 or 0. That level of rigor forces a different sort of thinking than anything you'd see in a natural language, and thus it needs to be taught very differently.
All of which is to say, I don't disagree with the notion that programming suffers from a problem in the way that it's being taught, but I do take issue with your over-generalized assertions regarding the lack of differences between spoken and programming languages.
* Pedantic caveat: The Venn diagram circles for "spoken languages" and "formal languages" can technically intersect if you do something silly like read C++ aloud. For the purposes of this discussion, I'm going to assume that "spoken languages" is a way of referring to everyday, spoken, natural languages, such as English or Chinese, and not a silly case like what I just mentioned.