Might be wrong but I thought Quake's software renderer does proper perspective texture mapping? (Original Half-Life is based on Quake)
While the FPU is busy doing that, the CPU can do the linear interpolation. Once the CPU is done with the 16 pixels, the FPU has also completed the divide and the CPU can do the next 16 pixels, and so on.
Looks pretty decent but is not 100% accurate.