How many rows of data is too many rows of data?

The magic number is billions. Until you get to billions of rows of data, you're not talking about very much data at all.

Do the math.

4-12 rows per user per course,... hundreds of courses and thousands of users?

400,000 to 1,200,000 rows. Let's assume 1000 bytes per row.

That's 400Mb to 1.2Gb of data. You can buy 100Gb drives for $299 at the Apple store. You can easily spend more than $299 of billable time sweating over details that don't much matter any more.

Until you get to 1Tb of data (1,000 Gb), you're not talking about much data at all.

I personally have had tables in production with 50 million rows, and this is small compared with I have heard. You might need to optimize your structure with partitioning but until you test your system in your environment you shouldn't waste time doing that. What you described is pretty small IMHO

I should add I was using SQL Server 2000 & 2005, each DBMS has its own sizing limitations.

100 (courses) * 1000 (users) * 10 (records) is only a million. That's the low end, but a decent database ought to handle it okay.

What sounds iffy are Name/Value pairs. That will limit your ability to correctly index things, which will be critical to good performance.