I agree. However, it can be hard to observe, even in the large, public sequencing data that is currently available. There is a large amount of mtDNA inserted into the nuclear genome, and this can make it appear that there is rampant heteroplasmy until you control for it. The actual contribution is extremely hard to observe.