One thing I’d separate in the results: are those downloads linked to the same people who played the demo, or are you dividing total downloads by total plays? If it’s the second, someone could download without playing, so 84.62% wouldn’t be a player-to-download conversion rate.
People who choose to launch a demo may already be more interested, too. A useful next comparison would be the same pack with and without the embed, keeping the price and page copy unchanged. I’d still value being able to inspect the animations; I’d just keep that benefit separate from what the numbers prove.