First, that's a paper from 2010. How was a paper from 2010 supposed to be "predicting" anything about what scientists in the past thought?
Secondly, and more importantly, I had been responding to Archangel Michael, who was talking about the thickness of the Antarctic ice sheet, not Antarctic sea ice. So your link about pack ice is totally irrelevant.
But hey, let's switch topics totally and talk about sea ice, since you seem to want to. Here's how the IPCC sums up all papers on the modelling of antarctic sea ice, including this one:
Whereas sea ice extent in the Arctic has decreased, sea ice extent in the Antarctic has very likely increased. Sea ice extent across the Southern Hemisphere over the year as a whole increased by 1.3– 1.67% per decade from 1979–2012 with the largest increase in the Ross Sea during the autumn, while sea ice extent decreased in the Amundsen-Bellingshausen Sea. The observed upward trend in Antarctic sea ice extent is found to be inconsistent with internal variability based on the residuals from a linear trend fitted to the observations, though this approach could underestimate multi-decadal variability. The CMIP5 simulations on average simulate a decrease in Antarctic sea ice extent , though Turner et al. (2013) find that approximately 10% of CMIP5 simulations exhibit an increasing trend in Antarctic sea ice extent larger than observed over the 1979-2005 period. However, Antarctic sea ice extent variability appears on average to be too large in the CMIP5 models . Overall, the shortness of the observed record and differences in simulated and observed variability preclude an assessment of whether or not the observed increase in Antarctic sea ice extent is inconsistent with internal variability. Based on Figure 10.16b and (Meehl et al., 2007b), the trend of Antarctic sea ice loss in simulations due to changes in forcing is weak (relative to the Arctic) and the internal variability is high, and thus the time necessary for detection is longer than in the Arctic.
Weak trend, short observed record, and high internal variability in the simulations. Which shouldn't be surprising, sea ice is a lot harder to model than ice sheet thickness, which really only has three main parameters - snowfall, melt/sublimation, and outflow, and the short observed record is due to how few people historically have navigated antarctic waters vs. arctic.
But again, to reiterate the primary point: the conversation you jumped into was about ice sheet thickness, not sea ice.