How does the optimizer decide whether or not to use parallelism when the parallelism threshold is specified in seconds (on *specific* hardware, see below) but the estimated CPU costs are supposedly unit-less?  Is the estimate treated as if it were seconds or is there a conversion happening internally based on some hardware metrics?

From MS:
[cost threshold for parallelism Option][1]

> The cost refers to an estimated elapsed time in seconds required to
> run the serial plan on a specific hardware configuration


[how-do-i-read-query-cost-and-is-it-always-a-percentage][2]

[sql-server-execution-plan-estimated-io-cost-estimated-cpu-cost-no-unit][3]

  [1]: http://technet.microsoft.com/en-us/library/ms188603%28v=sql.105%29.aspx
  [2]: https://dba.stackexchange.com/questions/9982/how-do-i-read-query-cost-and-is-it-always-a-percentage
  [3]: http://blog.sqlauthority.com/2010/01/22/sql-server-execution-plan-estimated-io-cost-estimated-cpu-cost-no-unit/