At least, for him to lose it at this point will be actually be a bit of a surprise. The numbers don't seem to be there for his opponent.
First, let's look at the total votes cast in Milwaukee and Outagamie counties (the two remaining to be canvassed), as of right now. Prosser's current vote margin: 7,310.
Assuming that there's simply no sudden appearance of 7,311 Kloppenburg ballots - and I invite any Democratic election fraud enthusiasts to try; I enjoy watching perp walks - the Democrats will need a net shift of those ballots for the same amount. That would mean a net 3 percentage point shift towards the Democrats... which is a lot, in this context. To put it another way: 7,310 votes is (in my opinion) outside of the margin of fraud.
As to turnout... the ?AP reported that the total combined votes for the 2011 Supreme Court race was 271,237 for the two counties, and 1,479,976 statewide. The 2010 Governor's election had a turnout of 406,276 for those two counties, and 2,160,832 statewide: in other words, voter turnout for the two counties for the 2011 Supreme Court race was 67% of turnout for the 2010 Governor's race... and statewide turnout in 2011 was 68% of 2010's. In other words: on first look there isn't anything obviously unusual about turnout in either county, in context; which is another argument against there being uncounted ballots that will appear in the re-canvassing*.
Bottom line? I'm going to be surprised if the vote totals change from the re-canvassing process. I recommend that the Democrats instead demand a very expensive and time-consuming recount, primarily because as it stands they will have to pay for it.
Moe Lane (crosspost)
*This is important because the discrepancy in Waukesha county's vote total was noticed at the time, which is one major reason why election fraud accusations are mostly coming from the fringe.