I'm the opposite of the GP, when I see the subjunctive not being used (e.g. "If I was a rich man"), it bugs me...and the proper use of the subjunctive is actually refreshing to me and is almost a litmus test for me that the writer really knows their grammar (and yes, I know the use of "their" will really rankle some people :) ).
That said, I don't think I would've ever been aware of the difference between "were" and "was" in the singular had I never studied a foreign language in high school, in which the subjunctive was explicitly defined.
If this is really an issue for you, you should probably be aware that there is no subjunctive form in "if I were a rich man" (CGOTEL terms that form "irrealis").
The english subjunctive can be observed in e.g. "I ask that you be there on time".
Why is there no subjunctive form in "If I were a rich man"? I do agree that "If I was a rich man" would also be technically correct in as many contexts, though, so that was a bad example (but I haven't had my coffee yet and that's the first example that comes to mind :) )
The first link is by one of the authors of CGOTEL, and comes out strongly against calling that were-form subjunctive; the second link is by another syntactician, and is a little more ambivalent.
What follows is a mixture of (a) my understanding of some terminology and facts, and (b) my interpretation of the arguments:
"Subjunctive" is a label for a type of sentence mood, where "mood" has to do with the relationship between the sentence and reality. Subjunctive mood is usually, like indicative and imperative moods, marked on the verb. (In my mind, interrogative is also a mood, but it is usually marked on the sentence, not on the verb.) Indicative and imperative moods have fairly narrow commonly-agreed semantics, but subjunctive is more of a catchall (my second link above will warn of the danger of confusing syntactic terminology with semantics, but while they don't correspond perfectly, they are related). We can quickly note that "if I were a rich man", a counterfactual clause, easily fits into area that can be covered by a "subjunctive" mood in various languages.
It's widely agreed that the following construction in English should be called "subjunctive":
I ask that you be polite to him.
The subjunctive form of "be" is so distinct from any other finite form that it makes for good examples. Any time a finite verb can be realized as "be", we're comfortable calling it subjunctive ("be he alive or be he dead, I'll grind his bones to make my bread").
That construction is productive today and nicely applies to any english verb:
I'm going to have to ask that your son return my necklace.
The argument against calling "if I were a rich man" subjunctive makes the following points:
- It doesn't use the subjunctive form.
- It is not obviously related to any construction that does use a subjunctive form.
- Though it is obviously a special form, it only applies to a single verb, and even there it's not required (Zwicky, my second link, talks about this some more). This can raise questions about how much it's a "real phenomenon", vs how much it's basically just an archaism.
- In the past, subjunctive forms could appear in English conditionals, but they no longer do.
Zwicky will tell you (and this is a point that follows my own inclinations well) that as long as you've described the rules correctly, the names that you give to the rules ("subjunctive") are irrelevant. So in that sense, you can call any form you want "subjunctive", as long as you're willing to extend the description to every instance of an analogous form, which is easy with "if I were a rich man", as it is the only form of its kind.
- However, as the generally-agreed subjunctive is a living and useful phenomenon, and is unrelated to the were-form, it might be nice to give the were-form a name that won't confuse people.
But we don't give the names of foreign languages' syntactic structures to any english sentence that happens to have the same semantics. If we're naming syntactic structures, then we want to divide sentences according to their syntax.
So, in Latin (and other languages, but I know Latin), there's a verb form called future tense:
laudo -> present 1sg active indicative, "I praise"
laudabo -> future 1sg active indicative, "I will praise"
English has nothing analogous. To achieve the same semantics (referring to the future), you need some combination of explicit time reference and periphrasis ("I will praise him tomorrow"). So we usually say that English has two tenses, which you might refer to as present and preterite, or past and non-past, or A and B, or any combination thereof (though I don't recommend "preterite and B"). Referring to a "future tense" in English is buying into the idea that every structure in every language has not only a conventional way of being expressed in every other language, but an analogous structure there as well.
On a totally unrelated note, the 1sg Latin subjunctive forms of esse [to be] are sim (present), essem (imperfect), fuerim （perfect), and fuissem (pluperfect). The Italian form "fossi" appears closely related to fuissem, so I'm a little surprised to see that it's called "imperfect". How many tenses of subjunctive does Italian have? Are there obvious analogs to any of the other three Latin forms?
See also: https://en.wikipedia.org/wiki/Subjunctive_mood#English