Powheg ggF_HZ process: add kappa_ghb settings and use nominal set of PDF.

A few updates to the Powheg ggF_HZ process that I think got lost in the new Powheg framework. Discovered while trying to update the ggZ(lep)H(bb) samples to the latest release.

• Add kappa_ghb, kappa_ght, kappa_ghz keywords, which are supported by the process for modifying Higgs couplings.
• Switch to the nominal PDF set (all NLO). Even though the diagrams are LO for ggHZ, the process is NLO in QCD as all diagrams contain a loop. This is what matters for the accuracy of the PDFs. Confirmed by @gmarchio (Higgs x-sec contact).

Tagging @cvarni @francav, who originally created the ggZ(lep)H(bb) samples. I'm not sure who to tag from PMG, so I'm including the MC JobOptions contacts: @cbecot @bouhova @joany

Also it is my understanding that release 21.6 is main one used by PMG. Please correct me if I am wrong.