Hi Joe,
Thanks for posting and for your interest in NCrunch.
None of the categories described in this Wikipedia article accurately describe NCrunch's coverage analysis.
NCrunch's code coverage is analysed line-by-line, not using language-specific constructs. This means that its code coverage is of a lower resolution than many other tools.
The reasoning behind this is that it is very difficult to perform high performance code coverage analysis and merging using higher resolution methods. NCrunch's needs are unique because it executes tests in real-time and in parallel, with potentially up to dozens (or even hundreds) of concurrent sessions running on many different machines. This means it needs to be very, very fast in processing and calculating coverage. The focus of NCrunch is that of a concurrent testing tool providing feedback as you type - it is not a specialised metrics analysis tool.
I hope this answers your question.
Cheers,
Remco