Peter wrote:
"What about the Bering sea? When I look at that Russian map, even the
"continental slope + 60 nm" line is, for the biggest part, nearer to
the shore than the 200 nm line (except the far western part of the
Donut Hole). So most of the Russian part of the Donut Hole is outside
of it. Does the Irish formula apply here then??"

To answer my own question: yes. At least acc. to the Russian deputy
minister for natural resources, defending the case at the UNCLOS
commission on the limits of the continental shelf:

"Within the area enclosed between the 200-nautical-mile exclusive
economic zone of the Russian Federation and the line delimiting the
maritime spaces of the Russian Federation and the United States of
America, the thickness of sedimentary cover (even within local arched
uplifts of the basement) everywhere exceeds 1 per cent of the
shortest distance from the FCS.
Thus, the data presented indicate that the part of the continental
shelf in the Bering Sea located beyond the limits of the 200-mile
zone from the baselines from which the breadth of the territorial sea
of the Russian Federation is measured and extending to the line of
delimitation of the maritime spaces of the Russian Federation and the
United States of America (USSR/United States Agreement of 1 June
1990) may be included in the continental shelf of the Russian
Federation, in full conformity with article 76, paragraph 4 (a) (i)
of the Convention."

(See http://www.un.org/Depts/los/clcs_new/documents/clcs31e.pdf )

"And is the seabed of an EEZ (normal 200 nm rule) always part of the
continental margin, regardless of these rules?"

Again, yes. See the UNCLOS, at
http://www.un.org/Depts/los/convention_agreements/texts/unclos/part6.h
tm , art. 76 (1).

Peter S.