The switch has X amount of RAM to receive and buffer frames from its ports. There is a risk of lost frames if a host/node transmits another frame (to the switch) and the switch cannot buffer that new frame because memory is full. This applies to switches that use store-and-forward.
See Switch Buffer Limitations in Network Switching Tutorial:
As packets are processed in the switch, they are held in buffers. If the destination segment is congested, the switch holds on to the packet as it waits for bandwidth to become available on the crowded segment. Buffers that are full present a problem. So some analysis of the buffer sizes and strategies for handling overflows is of interest for the technically inclined network designer.
In real world networks, crowded segments cause many problems, so their impact on switch consideration is not important for most users, since networks should be designed to eliminate crowded, congested segments. There are two strategies for handling full buffers. One is “backpressure flow control” which sends packets back upstream to the source nodes of packets that find a full buffer. This compares to the strategy of simply dropping the packet, and relying on the integrity features in networks to retransmit automatically. One solution spreads the problem in one segment to other segments, propagating the problem. The other solution causes retransmissions, and that resulting increase in load is not optimal. Neither strategy solves the problem, so switch vendors use large buffers and advise network managers to design switched network topologies to eliminate the source of the problem – congested segments.
Accord to LAN Switching, flow control could be employed to reduce frame loss.
Flow control is necessary when the destination port is receiving more traffic than it can handle. Since the buffers are only meant for absorbing peaks traffic, with excessive load frames may be dropped. It is a costly operation as delay is of the order of seconds for each dropped frame.
Traditional networks do not have a layer 2 flow control mechanism, and rely mainly on higher layers for this. Switches come with various flow control strategies depending on the vendors. Some switches upon finding that the destination port is overloaded will send jam message to the sender. Since the decoding of MAC address is fast and a switch can, in very little time, respond with a jam message, collision or packet loss can be avoided. To the sender, jam packet is like a virtual collision, so it will wait a random time before retransmitting. This strategy works as only those frames that go to the overloaded destination port are jammed and not the others.
Is the switch actually the traffic cop on that trunk for purposes of collisions and whose collision domain does this wire belong to?
Since it is whatever queuing mechanism that the switch employs (e.g. FIFO versus QoS) that determines the order of frames forwarded through each port, the switch can be regarded as a "traffic cop".
However this has nothing to do with collisions or collision domains.
Does this depend on using a dedicated WAN port vs using a normal switchport for the router connection?
Switches do not have WAN ports (although I have seen hubs with one port switchable for "normal" or "uplink").