Maybe this is what isn’t clear: the FIBRE network nodes do not only connect to mining pools. They are connected to each other and to mining pools in addition to making regular connections to random peers on the network. Any block received from any peer will be relayed as quickly as possible to all other FIBRE nodes and thereby forwarded as quickly as possible to all peers of each FIBRE node. Due to the careful placement of the FIBRE nodes to minimize latency between each other, this has the potential to reduce the collective latency of any block propagating through the network.
You can check that the provided source code does not selectively fail to relay or delay blocks of some origins, although it’s hard to prove that FIBRE actually runs the publicly provided code, of course.
Maybe this is what isn’t clear: the FIBRE network nodes do not only connect to mining pools. They are connected to each other and to mining pools in addition to making regular connections to random peers on the network. Any block received from any peer will be relayed as quickly as possible to all other FIBRE nodes and thereby forwarded as quickly as possible to all peers of each FIBRE node. Due to the careful placement of the FIBRE nodes to minimize latency between each other, this has the potential to reduce the collective latency of any block propagating through the network.
You can check that the provided source code does not selectively fail to relay or delay blocks of some origins, although it’s hard to prove that FIBRE actually runs the publicly provided code, of course.