In article , Leto2 wrote:
I was thinking about big lakes being created by the melting ice, and then
released, as happened in Canada, I think.

Certainly did, but whether it delayed the impact of the melt water on
the ocean by more than a few decades ... maybe for the American
mid-continental lake(s) (Lake Aggasiz?) but not so much generally.

I read also that this was one of the dangers of warming in the Andes : the
glaciers melt and get filled of water until they break down into an
avalanche. This happened above Huaraz, Peru, maybe 30 years ago. The town
disappeared.

That one was in response to a volcanic eruption under the ice cap on
top of IIRC Huarascan mountain. The timescale between eruption starting and
jokullhlaup was only a few days.
"Jokullhlaup" is an Icelandic term for the flood that comes from an
eruption under an icecap ("jokull"). There was an event about 6 or 7 years
ago when the Grimsvotn volcano erupted under (damn - forgot the name of the
jokull ; Google'll get it) and the flood came out about a week later.

Of course I suspect that at continent-scale, the slope cannot allow an
avalanche to happen.

This is a much more open question than you seem to think. There
certainly has been vigourous debate about the details in the past, and TTBOMK
there hasn't been a resolution. What is at issue is the stability (or lack of
stability) of the West Antarctic ice sheet. If it is possible for it to "go"
rapidly, then there's a potential 6 or 8 metres of global sea-level rise
almost overnight. It doesn't take much to see that it's worthwhile trying to
work out if this is going to/ could happen. It would seriously affect
flooding insurance premiums around the world.

--
Aidan Karley,
Aberdeen, Scotland,
Location: 57°10'11" N, 02°08'43" W (sub-tropical Aberdeen), 0.021233