What are the effects of running a stellar-core node on slow/weak hardware, e.g. not meeting stellar-core hw requirements?
Does that create a problem for consensus? If so, are only those affected, who use this slow node as validator?
If the node cannot catch up, it will fall behind in the ledger state. This is a problem for anyone who uses horizon on that node. However, it would only be a problem for consensus if other nodes placed the slow node in their quorum slices, which they are unlikely to do if you can't keep up with the network.