Google Trends is another data source to watch as this event approaches:
https://trends.google.com/trends/explore?geo=US&q=solar eclipse 2017
This shows the Wyoming-Oregon segment of the path having a greater interest, and potentially, traffic.
As for clouds, midday overcast here in the Midwest in August is fairly rare, outside of being in the trajectory of a daily MCS/ring-of-fire regime. Any cumulus fields *should* be in their infant stages during totality. Synoptically, I'll be concerned if a west-northwest flow regime sets up to advect MCS anvil material from the northern Plains/upper Midwest, or worse, if the jet is farther south and an overnight MCS is marching across eastern Kansas/Nebraska the night before. The most likely cloud problems would come from high-level anvil debris. Low-level stratus this time of year is extremely rare.
I'm almost resigned to the cloud issue due to the logistics of trying to reposition. If traffic levels are on the upper end of predictions, it is likely that repositioning any later than 36 hours out won't be possible. If it looks grim the night before (MCS in NE/KS heading east) I will consider it based on realtime traffic data.