The problem with this "fix" is that it misses the most fundamental of points here. The mere fact that some buyers will give a low star rating for shipping charges even when shipping is free simply proves that the star rating system doesn't work - buyers will give ratings based on all kinds of things which don't relate to what they are meant to be rating.
Example: a buyer over-bid and then suffers from bidder's remorse. Somehow, he/she has to find a scapegoat. So,the seller gets poor ratings out of the buyer's frustration at having spent too much, not because the seller did something wrong.
Or take shipping charges: we sell small and lightweight items. Often buyers complain because they think that something this small and light should practically cost nothing to ship - they ignore the reality that the post office has minimum charges, below which nothing can be shipped. Buyers tell us that "there must be a cheaper way", without ever telling us what it is - we get the blame for the post office charges.
Or shipping time: on average, we ship within 2 work hours of receiving payment - it simply can't be done any faster than that. Yet sometimes buyers give us low ratings because it took too long for the goods to arrive... like the buyer who gave us a 1 or 2 score because the goods had not arrived within 3 days of his making payment (we are located in Australia, which he knew, he is located in the US, yet somehow he expected a shipment from half-way around the world and which has pass through customs, to reach him in less than 3 or 4 days!
The point is this: the star ratings do not measure what they are meant to measure, and therefore they are less than useless, because they can be outright misleading.
No "automatic 5 stars for free shipping" can fix that basic flaw.
Quite apart from which it discriminates against buyers who sell lower-priced articles - if your margin on a product is $2, you can't give away $4 in free shipping!