It all comes down to the price of pushing bits. The big cloud providers have brought that cost so low, that you have to provide an incredible amount of value with a SAAS offering now for it to make financial sense anymore.
It's questionable as if spark every would've been nearly as successful of a product if they had done this.
Sparks success is that it came out of Berkeley, was better than traditional map reduce in a number of ways initially, but more than that it was able to spread because a lot of map reduce product providers were able to push it for their clients because of it's apache license explicitly.