Order of block retrieval

I’m trying to understand if IPFS qualifies as a good replacement for bittorrent when distributing files within a private network.

Say I want to share a file of some hundred MB to more than a thousand nodes. That file is initially seeded by one central node, and the other nodes are informed at virtually the same time to download it simultaneously. With bittorrent, each node would start to download a random block from the origin and by that distribute the load evenly.

If I understand correctly, this is different IPFS however. Due to its streaming capability, all nodes would read that file starting with its first block, and concurrent downloaders would race against each other, leading to a high probability that only the central node is able to seed. Is this correct?