Here are some key points about parasitoid wasp evolution:
They started radiating about 120 million years ago, at the same time as flowering plants... Today they represent 10-20% of all insect species.
They have accompanied the other insect species in their diversification since that time, and often undergo co-speciation... When a host species divides into different types, the wasp species divides too.
Species that lay inside living insects are thought to be derived from wasps that lay eggs on or in other insect's eggs or pupae. Endoparasitoids have evolved from ectoparasitoids independently in different hymenopteran lineages.
Some species also lay microeggs on leaves near a larvae, and when the host eats the leaf and the egg, it develops inside them.
They have both a sting and an ovipositor, which permits different species a variety of specialized and opportunistic behaviours to lay on or inside eggs, and to inject genes and venoms which inject into the host to modulate host immunity, metabolism and development. Up to 70 different venom proteins have been found in a single wasp species.
Parasitoids can also inject substances like polydnaviruses (PDVs), virus-like particles (VLPs), ovarian fluids and teratocytes.
This kindof reflects the complex evolution of the Pepsis species described... The different components of venoms and behaviours came as little changes over 100ds of millions of years, and the specialization on a single species of tarantula kindof gives the wasp exclusivity from intra clade competition.
Digging a hole to put the spider in does seem overly complex. apparently, it must have first learnt to pull the animal into a hidden recess... Afterwards it started to scratch and clear the recess, and that foraging movement eventually turned into excavation of a pit prior to dragging the prey.
here's one of the most readable references: https://www.nature.com/articles/srep19604