Thank you for your reply. I'm new to this and I'm not sure I fully understand, so please forgive my naive follow up questions.
I currently have only the seed node on a server and three additional small devices, each with a node. I've observed that even when a device is not actively publishing, the disk usage in the /home/user/.multichain/chain-name/wallet folder continues to grow. Is the device receiving the publishings of the other two devices? If so, do the calculations above require a multiplier of the number of nodes in the chain?
For item #3 in your reply, does each publication to a stream consume a block?
Is the information in the blockchain compressed in any way?
I'm thinking that if there is no compression, each publication to the stream uses a block and all nodes receive all publications from all other nodes, then in my original scenario of 10,000 noded publising 4 times every 15 seconds, in a single hour the disk usage per node would be 10,000 nodes x 4 publications x 4 time a minute x 60 minutes x (2048 byte block overhead + 200 transaction overhead + 400 transaction payload) = approx 25Gb.
Please tell me I'm grossly misunderstanding this scenario, as the devices would exhaust their respective capacities in less than an hour by my calculations.