Migrating to new materialization

[rlh1994]

The package is soon going to be tested for and then migrated to the new optimised incremental strategy. See the utils PR here: snowplow/dbt-snowplow-utils#117

In general the change is quite straight forward. For the models:

• Swap out the materialization in the config
• Add the snowplow_optimise setting in the config and set it to true
• Provide the incremental strategy (technically not required but adds benefit for bigquery and lets users now override this in a single place in a variable).

There are a few open questions around the best way to manage some of this still, which it would be good to get opinions on.

1. Do we want to still provide the snowplow__incremental_materialization in the package or remove this entirely and just hard-code the incremental method for those tables?
   2) While the incremental strategy does allow us to use the already optimised insert_overwrite for bigquery, it also adds an extra line to every config and an extra variable. In theory this method is the same across all incremental models where it is called, so the user could set it at a models level in their project if they wanted to. If we removed it and they did not set it, they still get the same benefit as our method is close to the way that method uses partitions anyway.
2. If snowplow__disable_upsert_lookback should be listed in the variables in the project or just in our docs site?

[rlh1994]

It turns out that the bigquery method replaces the entire partition and so expects the full partition to be in the temp table, which it isn't for us. Therefore, we need to use standard incremental for bigquery as well!