There is much code repetition in tests. Fix that

Make use of libraries where possible, make libraries out of low level operations, make verifier clean af.

Events are using cheap logdata space. In average, 600K extra gas is needed to send the whole proof+siblings as an event. Check if this can make the contract more practical in combination/replacing calldata model.

The metrics are measured with optimized contracts but let's commit that to master after the refactoring of the structure of tests

1. Refer to summa-tx (and possibly other projects) as know uses of the hash-and-resubmit pattern.
   I think the pattern is used here https://github.com/summa-tx/relays/blob/master/solidity/contracts/OnDemandSPV.sol#L320 - lines 340 and 357.
2. Acknowledgements to Patrick McCorry (refs for optimistic things) and James Prestwich to acknowledgements (h&r reference and fixes for EIP2028).
3. Add cost of submit in cryptoeconomics sections for different fork indices.
4. Check the cost of sha2 and sha3 in appendix.
5. Update the cost of load data from 68 to 16. This is due to EIP2028 (in current Istanbul fork) which has not been updated in the yellow paper yet.
6. Editorial pass on sections 4, 5.
7. Show that the contract is resilient to DOS attack. This can be derived from item 3 if we use a fork chain from genesis.
8. Add event-based proof acquisition in h&r variations. (optional)

gasleft() does not help and profiler does not display this. Try emiting of loging it.

• Delete = 0 s in variable initialization
• Replace uint with uint256
• Remove unneeded comments
• Add proper documentation
• Hide low level functions in separate lib

k is not used whatsoever and the usage of m is wrong. Let alone it is set to only 6.

High gas price increases the probability of a tx to be included in next blocks but there is no guarantee. For 10Gwei (highers price) the probability of being accepted by last 200 blocks is 90% which doesn't seem enough in case of high bets.

Remove applications and fix structure at thesis text