Due to the use of scientific notation in the definitions of default values for the number of permutations in these two commands, the floating-point version of template function get_option_value<>() was silently invoked, rather than the intended integer version. This meant that any value passed by the user at the command-line would be interpreted as a floating-point number. In cases where a decimal multiplier was specified (e.g. 1B = one billion), this would not be interpreted correctly, as such relies on parsing as an integer. Previously, this usage would silently result in a value of 1, since the trailing "B" would be silently ignored. The issue was detected due to #1794: such usage now throws an error since conversion of string "1B" to floating-point fails to consume the full string. This commit fixes the parsing of specifically these two command-line options.

Prior change in the impelementation of strong FWE control, changing the command-line option from receiving a boolean to being a flag, was not propagated to the point in code at which that option was read. This however now results in an error upon attempting to use that option due to merge of #1794, as it erroneously attempts to convert the next item on the command-line following the -strong option to a double.