add gp argument to epinow2 entry

[avallecam]

In https://epiverse-trace.github.io/howto/analyses/reconstruct_transmission/estimate_infections.html
Fix as suggested in epiforecasts/EpiNow2#989 (comment)

[jamesmbaazam]

I don't think this should be labelled as a bug but rather a modelling decision. Moreover, I think it would be better to leave the default setup that corresponds to the pinned version of the package instead of trying to preserve the behaviour from an older version of the package.