I hate the prospect of leaving a few percentage points of performance on the table for what is an otherwise simple change. Or am I just massively overthinking this and it's going to make no meaningful difference?

Unless you use negligible hot water the energy used by the compressor will dominate, the water pump will be negligible.  So you should optimise for that.  Some ASHPs have more than one option for DGHW (eg an 'eco' setting) which takes longer but uses less juice.  Does yours?

Yes, I already run in the Samsung equivalent Quiet Mode for DHW, which limits output to around 70% and which is more efficient.

Now I'm looking at the effects of flow rate. I agree the potential electrical savings for different circulation pump speeds is negligible at ~1% of overall electrical input.

I hate the prospect of leaving a few percentage points of performance on the table for what is an otherwise simple change. Or am I just massively overthinking this and it's going to make no meaningful difference?

Unless you use negligible hot water the energy used by the compressor will dominate, the water pump will be negligible.  So you should optimise for that.  Some ASHPs have more than one option for DGHW (eg an 'eco' setting) which takes longer but uses less juice.  Does yours?

My system uses fixed rate pumps, so is it more efficient to set a faster or slower flow rate for the DHW cycle?

A slower flow rate presumably allows more time for the heat in the circulating water to transfer into the DHW tank and thus widens the dT between flow and return, which means the final LWT would need to be higher to achieve any given DHW tank set point?

That would suggest a higher flow rate may be more cost effective? But then there's the additional electricity cost of running the circulation pump(s) faster, albeit for a relatively short period of time (my Wilo pumps use ~30Wh more running at max speed versus turning them down by one setting, so maybe 15W over the 30min DHW cycle so probably negligible at a saving of around 1% of the overall electrical usage).

I hate the prospect of leaving a few percentage points of performance on the table for what is an otherwise simple change. Or am I just massively overthinking this and it's going to make no meaningful difference?