Without see actual examples it is pretty hard to be specific. But if the goal is the CSV at the end I would suspect you could read Excel File #1 directly with SAS and then do the calculations and export a CSV with the results.
I basically did that for three files this morning.
The likely main concern is the actual structure of Excel #1 as there are some very badly designed formats of data in the spreadsheet world and reading them can be problematic. With a data set of a specific format created from the starting data then the manipulations done with your steps 3 and 4 are very likely much more stable in SAS (no copy and/or paste errors). One of the pretty nice things of working with SAS code is such a process is that when the next update to Excel comes out (potentially breaking all sorts of macros ) then the SAS code is likely to run. You'll find processes written in SAS such as SAS 5 from the 80's that the code still runs.
Hopefully someone in your organization actually can describe / understand what the existing macros are doing. Note: literal translation of VBA macros is generally suboptimal as SAS provides lots of manipulation tools that work a bit differently. Some things that may be called "interpolation" may be amenable to regression models (that may only need to be developed once) and scoring new data at specific values for instance.