I assume you are talking about millions/billions of years of seasonal snow/ice pack melt to form those gullies, much in the way glaciers here on earth formed gullies/valleys? Where did the *melt* go?

Yes, millions of years or repeated cycles of annual snow accumulation and thaw (in a seasonal melt model like Fracois Forget's, or my own CO2 model)

For Christensen's model, thousands to tens of thousands of years of accumulation then repeated pulses of melt for hundreds to thousands of years.

Not really like glaciers on Earth, since those move as solid ice and gouge their valleys.

Where did the melt go? - back into the atmosphere, and back to frost elsewhere on the planet - to form new snowpack next year in other alcoves and possible gullies, or to join back into the polar icecap for another few milleniia - the water cycle on Mars is long, slow, and weak compared to Earth, "but still it moves", to misquote Gallileo...

And yes, many craters have central peaks, with gullies flowing down them, although in this particular example I don't see good signs of any.

This is one of the major planks in the anti-groundwater argument, since that's the last place you expect groundwater to escape from (unless you make an extreme model that these are H2O "volcanoes" sourced from very deep underground and exploiting the zone of weakness of the central uplift of the crater. I haven't seen that model in the literature, yet, but it will come, I'm sure...)