Oh I knew people would say that. But it isn't about doing the right thing. I knew giving both would be morally right and the most correct solution academically speaking. This is the first time in my life I have ever considered even the slightest diversion from the most morally right option. But, I think as soon as you put the websites in, it looks less solid a piece of work. Doesn't anyone else think so?
I just wanted to offer a thought on this idea before we get too far from it:
In Biblical / Classical Studies (and fields drawing on that literature), the standard citation format is to give generic "book chapter verse" (or book-line or whatever) references in the paper, itself. For example, you would cite using something like Plato, Republic VII.522. Or Genesis 1:12.
The idea is that anyone using any original or translated text (with standardized annotations) would be able to locate your reference.
Often, if you quote, you may cite your translation (or indicate in footnotes you're using your own); if there's no translation citation for a quote, the general assumption is, I think, that the translation is the author's own.
In your bibliography, you would note the edition/publication of your text (also, it's not at all uncommon for people doing work in, say, classics to reference standard texts published online - particularly through the Packard Humanities Index texts).
If your work is using primary, archive documents, wouldn't that be analogus. Your citations would then be "Letter of January 12, 1889). You would then have a footnote indicating where these materials can be found (including the note: archive available online at xyz.org).
That's not at all dishonest.
I can't imagine how it could be viewed otherwise.
Now, if you make a big deal out of or make an argument suggesting you, physically, personally went to look at the archived item...