In engineering terms consider a new super dam that has been built. Suddenly the engineers realise the builders used weak cement and didn't properly install the foundations.
As it fills the dam starts to leak from the bottom and perhaps the middle. You know that it will go and that as the dam fills and the pressure rises the probability of an earlier collapse increases dramatically.
Predicting structural failure when it is imminent isn't that difficult since you know that it's about to happen. You have actual proof that the structure isn't sound since the effects of this are clearly visible.
But predicting something 200 to 1000 years into the future is way beyond the abilities of civil engineering. Sure, you can make a prediction based on what is known today but there is a huge risk that this will end up being inaccurate over that length of time. You only need a slight change in the rate of degradation, compounded over a very long time, to end up being wrong in your forecasts if they go that far into the future.
External factors are also a major influence. Eg the Tasman Bridge was opened in 1964. I'm not sure of the actual design life, but it would likely be a century or so for a bridge like that. Then a ship rammed into it in 1975 causing a partial collapse. Needless to say, whilst it was always possible for such an event to occur nobody could have accurately predicted the timing unless by chance.
Or take the coal mine at Yallourn (Vic). Circa 1920 they took samples and used these as the basis for designing the boilers at the first Yallourn power station (Yallourn A). The power station was built and the mine developed. Just one problem - the samples they took turned out to be different to the coal from the mine that was actually developed to the point that the coal wouldn't burn in the boilers. End result = various attempts to modify the boilers and the SECV had to operate a second mine close by, which had coal of the type in the samples, in order to make use of the power station they'd just built. A simple mistake but a real one that was actually made.
Now, can anyone tell me that we know
exactly all relevant information relating to this ice? No, you can't say that because there are uncertainties. And with uncertainties comes the possibility of error. And there will be external influences too - whilst it might be given that the ice has become detached and will eventually melt, the rate of melting would almost certainly change if the temperature changed. And there is a definite possibility that the water temperature won't be constant for the next 200 - 1000 years. Even without the man-made climate issue, natural variation alone is almost certain to result in a change over that time either up or down.
I am not an engineer but I know a few and none of them would be willing to make any definitive statements that far into the future about something as uncertain as naturally occurring ice without stating that they are estimates only, are based on current knowledge which is likely to be imperfect and are subject to ongoing revision as more data becomes available.