Hi Jamie, I have bad and good news.
bad news: currently there is a limitation that after you connect datasets using 1:N relation, you cannot set referenced attribute (foreign key reference in Services dataset in your case) as primary key of the dataset. Therefore the connection dialog suggests to create new “field” for such relation instead of using Service id attribute which is already present in :Services dataset. You can do it and map it to the same source column in the Services table. You will end up with two Service Id attributes in logical model (one contained in Service id lookup and other in Services dataset) as is currently the case in your model anyway, but the datasets will be connected so you will be able to use attributes from both in the same insight.
But I have also good news: a feature is coming which will let you use composite primary keys and composite key references. This feature will also let you set primary key consisting of reference to attribute in other dataset. It will be possible to set Service id column in Services dataset as reference to primary key attribute in Service id lookup and as primary key of Services dataset at the same time once this feature is delivered. Note that you are using two Service Id attributes so if you want to have only one in your data model, it will be better to connect datasets using the new field first (as is currently possible) then go through insights/metrics/dashboards which use Service Id attribute from Services dataset and replace them with Service id from Service lookup datasets before you remove the Service id attribute in Services dataset. You can edit title of the attribute in modeler to differentiate them better in data catalog.
Public release of the composite keys in GD CLoud is planned for the end of November, but you may contact support to have a early access after release which is planned for this Thursday. Watch
release notes.