More research will be done on the expectation values for the
likelihood statistic. This will include several possible situations
(purely poisonian coincidences, real correlations and mixtures).
This will help distinguishing reliable corrections from unreliable corrections.
This implies that some additional keywords may be added in future
which give information about the correlation.
Some research could be done to other methods for determining offset.
One can think of Fast Fourier Transforms for fast correlations, or a method
based on triangle patterns.