In your scenario I would create a single cube that includes all companies, but use perspectives to create "cube views" for each separate company. The drawback here would be that...

Please post your MDX query along with your results and what the expected results are.

Try the rank() MDX function. Reference available here: https://docs.microsoft.com/en-us/sql/mdx/rank-mdx

I wouldn't call it a "mistake" as the instructor did, but I do like to pad source strings a little to account for future growth. Source applications can sometimes change...

Unless you have implemented row-level security, I don't think it is an impersonation or permissions issue. I also don't think it has anything to do with your Git repository. 

A few questions: Is your cube hosted on a server inside your network (i.e. behind your firewall)? If it is, your external clients will not be able to get to...

In that case your first example is the only way to achieve that. I was not able to find any other plausible way.

Cannot necessarily comment on the good/bad aspect of the above, but I would be inclined to use named sets. See reference here: https://technet.microsoft.com/en-us/library/ms145487(v=sql.110).aspx

You could create calculated members for all of the dimensional attributes, or perform the union in SSRS (assuming that is what your reporting platform is).

As you stated above, the query will return duplicates because of the overlaps in granularity. I'd recommend one of the following approaches:

• Build the MDX dynamically based on the user selection,...

Yes...if you're using 2-factor authentication in Azure, you need to set up an app password as Power Pivot may not support 2-factor authentication natively. See the following: https://docs.microsoft.com/en-us/azure/multi-factor-authentication/end-user/multi-factor-authentication-end-user-app-passwords

Look at the Lag() MDX function here: https://msdn.microsoft.com/en-us/library/ms144866.aspx

Using the above function in conjunction with the .CurrentMember property, you should be able to achieve what you need.