Bouncing board needed for use of Power Platform Dataflows

[D_PBI]

Hi,
I would like to some guidance/confirmation on the use of Power Platform Dataflows. I'll try and keep this short.
I work for a company that is MS cloud based with our core solution being Dynamics 365. Any third-party data is provided via Excel files. As our third-party partners are growing, it's now time to introduce an efficient process to automate data entering our domain.
At a high level, I looked into the use of Azure Synapse, Data Factory, etc.. but considering the small datasets we need from our third-parties and our current expertise in technologies, I feel the use of Power Platform (not PBI) Dataflows into a DataVerse is an appropriate route. I've seen demos online but I have some questions.

1) For the very first data load from a third-party data source you would choose 'Load to new table' so that data source table/view has its target DataVerse table generated for the data to be loaded to. After this initial load you would no longer use this particular Dataflow but instead create a new Dataflow choosing the 'Load to existing table' option. It would be against this latter Dataflow that you'd schedule a data refresh, and delete the initial Dataflow (the 'Load to new table' one). Would this assumption be correct?

2) Our Dynamics 365 solution has a UAT, Pre-Prod and Production environment and so we have that many DataVerse environment domains as you can only have one or zero DataVerse database per environment. I would only look to bring the third-party data into a single/Production environment. To save on costs, would it be appropriate to load the third-party data into the Production DataVerse environment (the same environment as our Production Dynamics 365 database lives) using newly created tables, or would you suggest to create another DataVerse environment domain solely for third-party data?  If yes, why?

3) The third-party data comes from their transactional system. The data provided in the Excel files is current data. I don't believe for our basic use we need anything other than current data. However, should we decide we want to capture historical changes and keep history of individual records, is the incremental load process only a Premium Per User feature?

I won't know until I attempt to bring altogether what hurdles I'll face but my intention is to extract the relevant third-party data (so it will likely be a dozen columns, few hundred thousand rows deep) from a source table/view and load it into a staging table so our domain holds a raw, unchanged copy of the table/view. This staging table can be used by other departments to load raw data to, say, our Dynamics 365 solution, or it can be used directly in our Power BI reporting, or be merged with a different dataset via another Dataflow into a downstream table acting as a DataMart. This is my intention.

Is anyone able to confirm the above is accurate or guide me on anything I may have not considered?
If there is a informative video online that guides one step-by-step, it'll be a great help if it can be shared with me?
Thanking you in advance.