refactor EmptyBlock generation due to IBC support

[mtsitrin]

due to https://github.com/dymensionxyz/research/issues/173:

relayer both are expecting at least one more block to be produced after the ibc-transfer tx block before dispatching the ibc-transfer.

we've implemented mechanism that tries to generate empty block on submission

// We try and produce an empty block to make sure relevant ibc messages will pass through during the batch submission: https://github.com/dymensionxyz/research/issues/173.

// Verify the last block in the batch is an empty block and that no ibc messages has accidentially passed through.
// This block may not be empty if another block has passed it in line. If that's the case our empty block request will
// be sent to the next batch.
if !isLastBlockEmpty {
	m.logger.Info("Last block in batch is not an empty block. Requesting for an empty block creation", "endHeight", actualEndHeight)
		m.produceEmptyBlockCh <- true
	}

This mechanism can be improved by the following:

• RDK ante handler will mark IBC transfer packets with defined priority/code (ResponseCheckTx)
• dymint will treat this field to force next block production.

this improves the mechanism, as it's not looks for empty blocks, but just validates that we have a block post ibc transfer

[omritoptix]

I think it's a good optimization to have but sounds like a lot of work for now. I think there are lower hanging fruits which relates to either UX or Usability. Lets prioritize it last for the current refactor sprint.

[danwt]

Can we close it? I doubt we'll do it Reopen if you feel different