The replica should be able to handle the same rate of writes as its source, or else it may fall behind. This is called replication lag.
The replica may get a chance to catch up if the writes on the source come in bursts. Even if the replica falls behind, it can catch up during periods of low traffic.
But I've seen cases where the source is handling a more or less constant stream of writes, so the replica falls behind further and further, until it can't catch up.
Also consider that on the source, writes may occur in many threads concurrently. Traditionally MySQL replays those changes in a single thread, which makes it even harder for it to keep up with heavy traffic.
Recent versions of MySQL optionally can execute changes on the replica in multiple threads (but still probably fewer threads than the changes executed on the source).
Also consider failover. If you intend that the replica will become the main instance if the source needs to go down, then a replica server with half power may not be able to handle the traffic that your source server can.