Firstly I could easily see how FPTP would change the voters that mattered-obviously this would depend a great deal on the boundary. similarly I could easily see how it could change the balance between the large parties even if they all remained represented.
Secondly from the examples he gives it does sound as if some of the smaller parties such as National Union might be eliminated from the Knesset (though it does sound as if the effect might be surprisingly small).
Thirdly the small broad protest parties that pop up from now and again whether the Green's or the Pensioners Party would be in trouble .
Fourthly though and most fundamentally I just do not think that the party system would remain the same for long under PR-it's possible for big parties to dominate under PR-that is more or less what happens in Sweden (particularly if one remembers that the center, Liberals and christian democrats are closely aligned with the moderates/ conservatives) and the left and Greens only marginally less with the Social Democrats) . However what First Past the Post does (outside huge concentration of votes which appears to exist in Israel only in Arab and ultra-orthodox areas) is weaken small parties. It also provides an electoral incentive to combine before the election rather than after. Israel does indeed have a huge number of ideologically charged cleavages Jew vs Arab , secular vs traditionalist, hawk vs dove and so forth. Crucially also these cleavages overlap-that is there are secular ardent hawks and traditionalist etc. However the advantages that say Likud and Yisrael Beiteinu would have gained through merging (even if both lost a substantial strand of their support) in a FTPT system would create an enormous incentive to merge parties(this is even more obvious for labor and Kadima)
Similarly the risk a vote would be "wasted" would probably cause tactical voting and a leaching of votes whether secular dovish Ashkenazim from Meretz to Kadima or (now less likely labor) or hawkish traditionalist Sephardi from Shas to Likud.
Thus it strikes me just looking at the raw vote totals is a poor guide. One has to consider how the increased chance of a "wasted" vote would affect voter behaviour and the changed incentives for political parties to merge/ split. It seems to me this would have a big effect on political behaviour and then the tendency of FPTP to be rigged in favour of the winners would take effect.
IN other words what matters is not just the direct effect-but the incentives for voters and parties. These strike me as central to examing what the effect would be.