Consensus Cleanup: demo of slow blocks on Signet

[ajtowns]

@AntoineP

@ajtowns is going to mine the six blocks in a row and then reorg them out, so that anyone running a Signet node at the time would see them, but they would not impose a cost to IBD forever.

For those interested in the behind the scenes (including, perhaps, future me; hi future me!) what I did was:

1. stop the miner
2. make sure the next scheduled block was 10+ minutes away so the backup miner wouldn’t kick in too soon
3. copy @AntoineP’s proposed blocks into the custom-blocks dir targetting the next block heights (see inq#111 for the code to support custom blocks)
4. use just the core node to manually mine the next 6-8 blocks; 6 blocks of attack txs, then a couple of extra blocks with some tx spam so the mempool has something to do during reorg
5. on the core node, run invalidateblock on the first attack block
6. restart the normal miner, so that 7-9 blocks get mined over the course of about an hour to reorg the attack blocks out
7. wait for my super-slow node to finish validating all the blocks (at ~5m per block, yikes. on the upside, great for catching performance regressions in normal usage!)
8. run invalidateblock on inquisition.bitcoin-signet.net to ensure the reorg chain has good connectivity