Russian characters in the address(site.ru/статья.html) — possible problems?

The main one problem, and it is, as always, Internet Explorer that displays the URL in the address bar in decoded only when you manually set it or when you click on an unencrypted (and, therefore, it is incorrect according to RFC 3986) link.

Cm. bug report (vote for a fix is impossible, but you can click on the link "User(s) can reproduce this bug → I can too").

References of the form
\rru.wikipedia.org/wiki/%D0%A0%D0%BE%D1%81%D1%81%D0%B8%D1%8F
\r
And so I have Russian and all are indexed without any problems:)

urlencoding+UTF-8
\rread more

For me, the main drawback is if you copy the SEF URL from the address bar to the clipboard we are not SEF (encoded) URL, and much longer length (one character Cyrillic -> 6 characters of the URL), so I prefer to use the transliteration (system B GOST 7.79-2000, aka ISO 9-95)

If you mean these URLs:
\r
\rit-tales.ru/os/linux/настройка-ca-на-linux.html
\r
... with them, there are nuances, both positive and negative.
\r
Positive:
\r
1. Address normal read (unlike nastrojka-visokoproizvoditelnogo-servera.html).
2. The major search engines highly rank matches your search phrase with the URL (example on the same page see here)
\r
Negative:
\r
1. In the web browser or when copypaste address will look completely unreadable.
2. Some search engines are not compatible with such URLs and/or no index.

And you see how many pages are in the results of search engines abracadabras in the address links especially from search pages on the website and all is fine indexed and working.

Works fine. Tested on WordPress'e when doing articles directory.