This is generally due to job queueing on the Salesforce side, especially since it's stuck in a pending state. I can confirm that to be the case based on the provided log, e.g.:
2022-05-17 12:29:25,594 [ 215039] FINE - #c.i.i.b.ForceComMetadataRetriever - Submitted retrieve request with async process ID 09S25000004E9WmEAK
2022-05-17 12:29:25,595 [ 215040] FINE - #c.i.i.b.ForceComMetadataRetriever - Polling the status of the retrieve request.
2022-05-17 12:29:25,782 [ 215227] FINE - #c.i.i.b.ForceComMetadataRetriever - Status: Pending
2022-05-17 12:29:25,783 [ 215228] FINE - #c.i.i.b.ForceComMetadataRetriever - Retrieve request is still processing.
2022-05-17 12:29:25,783 [ 215228] FINE - #c.i.u.VariableLengthPollingInterval - ForceComMetadataRetriever.retrieve: Using polling interval 1000 ms for polling iteration 1.
2022-05-17 12:29:26,790 [ 216235] FINE - #c.i.i.b.ForceComMetadataRetriever - Polling the status of the retrieve request.
2022-05-17 12:29:26,990 [ 216435] FINE - #c.i.i.b.ForceComMetadataRetriever - Status: Pending
2022-05-17 12:29:26,990 [ 216435] FINE - #c.i.i.b.ForceComMetadataRetriever - Retrieve request is still processing.
2022-05-17 12:29:26,990 [ 216435] FINE - #c.i.u.VariableLengthPollingInterval - ForceComMetadataRetriever.retrieve: Using polling interval 1000 ms for polling iteration 2.
...
2022-05-17 13:42:53,806 [4623251] FINE - #c.i.u.VariableLengthPollingInterval - ForceComMetadataRetriever.retrieve: Using polling interval 5000 ms for polling iteration 791.
2022-05-17 13:42:58,814 [4628259] FINE - #c.i.i.b.ForceComMetadataRetriever - Polling the status of the retrieve request.
2022-05-17 13:42:59,358 [4628803] FINE - #c.i.i.b.ForceComMetadataRetriever - Status: Pending
2022-05-17 13:42:59,358 [4628803] FINE - #c.i.i.b.ForceComMetadataRetriever - Retrieve request is still processing.
2022-05-17 13:42:59,359 [4628804] FINE - #c.i.u.VariableLengthPollingInterval - ForceComMetadataRetriever.retrieve: Using polling interval 5000 ms for polling iteration 792.
Unfortunately this isn't something that IC2 (or any other tool) can control as it's completely based on server queue lengths. You would need to log a case with Salesforce as to why these queues have gotten worse recently.
Salesforce has now posted an official statement regarding this performance degradation:
https://help.salesforce.com/s/articleView?language=en_US&type=1&id=000365527