It is my viewpoint. There is nothing correct or not correct. But you can gauge yourself based on the flow, infrastructure, demography , homogeneous, heterogeneous sources etc etc.
By definition of DWH:
Subject-Oriented,Integrated,Time-Variant,Non-volatile.Also it is my trademark "Technology should serve business.Business should not bow down to technology" :)
If I, a salesperson, am transferred from one country to another by an organization.The question in
BI: sales data for me with my first country for data before the transfer date, and sales data for me in the new region after the transfer date. My identification in the same company remains the same. So sales dimension !!! A surrogate key allows me the same salesperson to participate in different locations in the dimension hierarchy.
Natural keys may be suitable when you have small, rarely changing, unique values (such as lookup tables), while surrogate keys are desirable for large, highly changing data key columns.
Also, I am pasting from design doc about surrogate keys:
Situations occur where the identification and choice of a simple primary key is difficult, if not
impossible. There might be no single column that uniquely identifies the tuples of a relation
variable or, looking ahead to physical design, there might be performance or query condition
considerations that argue against using a composite key. In these situations, and only in these
situations, surrogate keys are an ideal solution.
Composite key for load, if you use where conditions, maybe more time-consuming.