There were recently issues reported related to the hardforks being applied
**before** the actual hardfork in the mainnet when using the converter.
After some investigation it turned out to be a misunderstanding in how
the hardforks are actually applied - they are applied **after** applying
the actual block. That was not the case in the converter where hardforks
were applied just after receiveing a block. After trying to move the
hardfork checks to the end of the `convert_signed_block` function,
there was still a problem with the initial hardfork application where
it was harder to distinguish between "standard blocks" and "hardfork blocks".
Considering that the `check_for_hardfork` function is private withing
its scope and it is used only in one place, changing the block number
comparison from 'equal to or greater than' to just 'greater than' saves
a lot of redundant code.