Current sound synthesis techniques are computationally and monetarily expensive, and harmonically identical. The method presented above for organ sound synthesis produces functions that can describe unique waveforms at arbitrary frequencies within the range of the recordings. Since many parts of pipe speech are not modeled and other harmonic problems remain, this method is not fit for use in production systems. However, the results show a working proof-of-concept implementation, and suggest further work is warranted.
The author has implementations available online (http://mattjibson.com/schalmei/).