The node "type" to just store the blockchain and provide blocks (including witnessing, which what blockchains call to "block verification" and "production", aka signing the next block with new transactions that need to be included), can be ran on a Raspberry Pi yes... I have one Orange Pi (16GB RAM), and you could even run it on a 8GB thing. All it needs is space to store the entire blockchain (if one wants to, because you can also now store just part of the blockchain), which is around 600GB, so a 1TB SSD will do for a big while.

But what I completed here, is a different beast. It includes extra stuff (I will try to bring a post about it) in order for me (if I want) to point "ANY" app into my own server and not need any other node to interact with those applications that allow for point to custom nodes.

In summary, with what we call a "witness node" that runs on a Raspberry Pi, you would not be able to do this (for most apps, but some that just need to get blocks will work). With an API Node, you can.

For example, this is me, interacting with BeeSwap, entirely using my New Zealand infrastructure. Both Hive and Hive Engine APIs.

(note, that people would need to add my nodes as they are not there by default)

Nice suggestions. Will try write something up and get feedback.