Becoming cautiously optimistic for widespread Sunday rain in eastern lower MI

Two more model runs have come in since my original 50/50 sitting-on-the-fence skepticism post, and I'm delighted to say that the results are looking good for rain in the eastern portion of the state. The high-res NAM model in particular has put out some really nice indicators of just the type of soaking we'll need, though there's still some question as to how much coverage areas will see. If storm mode is discrete, as what's currently indicated, some areas may see very little to even no rain. Others may see a ton of rain. Here're some graphics to give you an idea:

This is taken from the 18z (2PM EDT) run of the high-res NAM4KM. As you can see, it predicts widespread rain to the tune of 2"+ over a great portion of lower Michigan. However, it's predicting a discrete storm mode where some counties and areas of counties will see much more than others. In a linear storm mode, coverage has better quality thanks to no gaps between storms. And while it remains to be seen if we'll be dealing with a discrete mode, linear mode or no mode at all (praying not), the 12z (8AM EDT) run has even more discrepancies, and higher totals over some areas due to it predicting some "training" of thunderstorms:

Huron, Tuscola, and Sanilac counties in particular would get a ridiculous amount of rain from this particular forecast, where as the rain would be much more limited to the southeastern portions of the state. Of course, this run can't see past its maximum distance, which means those areas still could get some decent rain coverage, and this is likely a massive exaggeration by the model, but you get the idea.

The other side of the coin of this forecast is storm severity. At this point I'm actually thinking the slight risk may increase in coverage over a wider portion of southeast Michigan, and some of the area might actually be looking at more of a widespread severe weather event should these forecasts verify. The storm mode these two runs have put out indicate discrete severe multicell clusters or discrete supercell thunderstorms, and some areas in particular would be looking at some enhanced potential for such storms. The storms themselves all appear to be right-moving in nature, which is of some concern. I'm not so sure on the large hail risk anymore due to some of the mid-level lapse rates looking a bit sheepish as compared to earlier forecasts, but there is definitely a damaging wind risk being forecast here. Tornado risk is still looking fairly low, but definitely non-zero, and storms that are discrete in nature have a greater chance of spinning up than do storms of a linear nature. Lake breezes are also a wild card in regards to turning non-tornadic storms into tornadic ones. I do not see much of a risk for significant tornadoes at this point, as wind profiles still remain forecast to be on the modest side and shear appears somewhat unidirectional, and even backveering in places. 

Overall, I'm still on the fence. The cap could still hold, and we could wind up with a whole-lotta-nothing. However, I'm definitely leaning more toward storm initiation than against at this point - about 60/40 in confidence level. The severe risk looks legit, but it'll definitely be worth it to get some much needed rain into the soil and onto the trees, which are quickly becoming dry kindling.

We'll get a much clearer picture by tomorrow throughout the day.