You have some very valid point Funky, let me try to help with a few of them:
Let me start by asking about a very confusing issue, to me anyway..... Now what I can't understand is: 1/ Where does the melt water come from? If its on the surface then how does it manage to seep anywhere as it would instantly freeze as soon as it left the surface, so how can it retain its heat for long enough or until its seeped to a level where it can continue to form a crack? This same argument does not appear to apply to the glaciers around Mt Everest where daytime temperatures far exceed anything recorded in the Arctic\Antarctic but where the ice remains frozen.
Melt water can seep a long way before becoming frozen again, it does form on the surface and does drip down into crevasses. If the ice is at nearly freezing, it requires little extra energy to melt it. In the summer in Antarctica, many glacier crevasses have a pool of near-freezing water at the bottom - it used to scare me to death thinking about it when walking across a glacier. A n icy lake at the bottom of a crevasse seems a lot worse to me than a dry crevasse. The same applies whereever the glacier is.
How crevassed an ice-berg is depends on where it came from and how it was formed. Some glaciers meet the sea at an angle that causes much crevassing, it is this and the terrain they flowed over that determines how many cracks they have and how deep they go.
Many Antarctic icebergs however come from iceshelves. Here the ice from the land meets the sea at a shallow angle, the land-sea boundary can come 10's even 100's of miles from the edge of the ice-shelf. Again depending on circumstances, the ice of the iceshelf will become compressed and consolidated again, so ice-shelf berg may not be so crevassed and readily broken up.
You have to appreciate that it's not always the average temperature that is the key factor to melting ice, but how many days the temperature is above freezing. An annual average of 5 below freezing with no days ever above freezing will mean no melt. An annual average of 10 below freezing with 10 days per year above freezing means a lot of melt.
"Back-yard science" - imagine you build a snow-man, it stays at 2 degrees below freezing for a month - snow-man lasts a month, but just one day with temperatures 5-10 above freezing which is perfectly possible before temps go back to -2, means at least a partly melted snow-man, if not a puddle.
The incident you refer to of water seeping into cracks in the Larsen B iceshelf is an example of a "tilting point". For thousands of years, there wasn't enough summer melt to cause the ice-shelf disintigration, but just like a tilting see-saw, eventually it tipped to the other position. While the underlying cause is a very small change, the effect is enormous because of exactly where it happened.
What you say about water vapour is true, it is rarely mentioned as its influence is very complicated. The fact is though that global temperatures are rising despite the effect of water vapour. The main difficulty I think is in understanding this is how a small overall planetry temperature of a degree or so can have such a large effect when we see daily and yearly temperatures fluctuate by far more than this.
The changes are seen at the "edges" first, like the nearly melted snow-man. If it's always -30 and goes to -28, that glacier overhanging your house is irrelevant. If it's always -1 and goes to +1, that glacier changes its character.