Some thing to consider is that, in the biofuel article you posted, It said the there was little enzymatic degradation from sonication, but I think you should consider that in their process, they are adding enzyme, most likely a purified and concentrated enzyme used in industrial processing, where there can be 1 - 5 fold excess of enzyme added to their process. But in your mash you will have the enzymes from the grain, so your concentration of enzymes will be much lower. Therefor any loss of enzymatic activity due to sonication will be much more evident in the mash, than in a Industrial process.
If the enzyme they are using is in excess, then the loss of enzyme not be reducing the enzyme below a threshold of activity, that would result in a observable difference in their process, so even with sonication they still reach the enzymatic steady state, of 3 fold higher efficiency.
In other words, if you need 100 units to process 300 grams of starch, and you are adding 500 units, but the sonication reduces your viable enzyme by 80% you will still have 100 units to process the 300 grams, so you might not see the degradation unless you are looking for it, in a change of rate.
My suggestion would be to try short to long periods of sonication, and measure your efficiency, but also be aware of off flavors that might get released due to sonication.
I use sonication in the lab quite a bit, and it is usually to shear DNA and proteins, but the frequencies are much higher and the volumes are much smaller. Sounds like an interesting experiment though.