Faster Horner scheme evaluation

[David-Berghaus]

@fredrik-johansson mentioned at the end of his paper "Faster arbitrary-precision dot product and matrix multiplication" that similar optimizations should be possible for the Horner scheme. I personally think that this would be a very useful feature to have.

[fredrik-johansson]

Yes, indeed. I will need to find a good block of time this year to write the code, it's a lot of work to do with a similar level of optimization.

[David-Berghaus]

That would be awesome, thank you! Yes, I thought about implementing it myself but it seems to be too technical for me...