fix!: reject zero Electrum batch sizes and Esplora concurrency

[reez]

Reject Electrum batch_size = 0 and Esplora parallel_requests = 0 before consuming FullScanRequest or SyncRequest, because these values currently return success without scanning the requested data and leave the request unusable.

[j-kon]

Hi @reez, I'd like to work on this if nobody else is already handling it.

From what I can see, both the Esplora and Electrum scan/sync paths currently consume the request with .take() before validating parallel_requests / batch_size.

My plan would be to validate the zero value first, return a binding-visible error without consuming the request, and add regression coverage that verifies the same request is still usable after the invalid call.

I'd cover:

• Esplora full_scan with parallel_requests = 0
• Esplora sync with parallel_requests = 0
• Electrum full_scan with batch_size = 0
• Electrum sync with batch_size = 0

Is that the intended scope? If so, happy to take it.