I still only foresee rain except the very northern edge of the precip. For lower MI there may be a brief mix at first before it becomes all rain, and stays that way. Maybe portions of northern lower MI may see snow, but no way in southern MI. Way too much warm air wrapping up in this storm.
I generally agree that it will be rain, but you just can't deny that if precipitation rates become heavy, the rapid melting of heavy snowfall (aloft) might cool the layer to isothermal 0C. We aren't too far off from 0C anyway (-3C), but the layer is rather thick... I thought this would happen last event too, but it didn't (precipitation was mostly light). FWIW... The latest NCEP WRF suggests precipitation rates of up to 0.75 to 1.00 per three hours for two timesteps (mostly elevated convective precipitation). That's alot more QPF than the other models were indicating, but I believe it is trying to depict convection.
I have been reviewing a few case studies where such events have occured. Look here, particularly at page 8:
http://www.spc.noaa.gov/publications/goss/kaingoss.pdf
It's a very slim chance and I certainly wouldn't be forecasting snow, but I won't completely rule it out.
Anyway, for other areas... I pulled up some soundings for Central and Eastern WI (ISW, GRB)... They are primarily AOB 0C all the way up. Looks like a SN/IP sounding to me versus ice, but then again QPF isn't really all that great up there. MKE has some good QPF amounts on the NAM, and it's profile is similar to areas east of there (GRR, DTX) with that pesky 2C-3C warm layer just above 850MB... SFC temps are a tough call, but it will be daytime and the sun would have an adverse affect on ice accumulations.
The 12Z run of the NAM is actually slightly warmer than it's 06Z run in the low levels. Interestingly, the 12Z NAM has slightly lowered QPF amounts, which may have a relationship to the temperature profile (even if it didn't drop the profile to 0C, it would still have SOME affect). I'm not sure if models can or even try to resolve dynamic cooling...