Here is my quick review to Welzl's draft on MulTFRC.
The abstract of the draft has a rather weak motivation. I think the abstract should be strenghtened, so as to make a stronger case for the need of the draft. Is the motivation to provide a mechanism to tune the aggressiveness of TFRC?
New equation X_Bps: Some explanation so as to why or at least a pointer to a paper that describes the equation derivation and properties is in order.The pointer provided [Dam+08] presents the results, and points to a tech report for derivation.The important point, however, is to have an idea of the range of rates produced by the equation. The original TFRC equation is much easier to understand and make a "safety" analysis of Bps possible range values than the proposed new equation IMHO.
An implementation consideration session is also helpful to discuss issues such as how to work around the lack of floating point operations in Linux Kernel, for instance...