One path to consider along these lines is the concept of "functional pseudocode" ... an intermediate language that can be translated into native idioms for whatever programming language you want.

If you are familiar with "Esperanto" then you have a general idea by what I mean when i say "functional pseudocode." If you are not familiar with Esperanto, it is a human-invented language that is intended to be an easy 'intermediate' language that anyone can learn, and translate into their own 'native' language.

I have toyed with this kind of thing in the programming language arena, and it sounds similar to what you are talking about. In theory, it sounds like a promising idea, but as you probably have noticed, there are some practical limitations.

Perhaps the biggest limitation is that if you want to invent 'functional pseudocode' what you are doing, in effect, is inventing your own programming language, and people can be very parochial when it comes to what programming languages they use, and in what circumstances (if they know more than one) ... it's the same with human language.

Having said that, there are some advantages to this kind of approach. Just as with human languages, your horizons and understanding can expand considerably if you become 'fluent' in more than one. The thing I like about perl is that it is amenable to many many different 'dialects' ... and thus is a good starting point if you want to explore this kind of thing further.