No poly sqrt velu

[GiacomoPope]

velu/ell=103            time:   [59.892 µs 59.967 µs 60.036 µs]
                        change: [-0.9363% -0.7370% -0.5721%] (p = 0.00 < 0.05)
                        Change within noise threshold.
velu/ell=211            time:   [120.48 µs 120.64 µs 120.80 µs]
                        change: [-0.8911% -0.7433% -0.5936%] (p = 0.00 < 0.05)
                        Change within noise threshold.
Found 1 outliers among 100 measurements (1.00%)
  1 (1.00%) high mild
velu/ell=311            time:   [177.33 µs 177.53 µs 177.73 µs]
                        change: [-0.7201% -0.5890% -0.4640%] (p = 0.00 < 0.05)
                        Change within noise threshold.
Found 3 outliers among 100 measurements (3.00%)
  3 (3.00%) low mild
velu/ell=587            time:   [334.65 µs 335.02 µs 335.39 µs]
                        change: [-0.7155% -0.5783% -0.4463%] (p = 0.00 < 0.05)
                        Change within noise threshold.

sqrt_velu/ell=103       time:   [59.030 µs 59.083 µs 59.132 µs]
                        change: [-26.322% -26.217% -26.107%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 10 outliers among 100 measurements (10.00%)
  9 (9.00%) low mild
  1 (1.00%) high severe
sqrt_velu/ell=211       time:   [90.040 µs 90.132 µs 90.220 µs]
                        change: [-25.769% -25.647% -25.522%] (p = 0.00 < 0.05)
                        Performance has improved.
sqrt_velu/ell=311       time:   [118.01 µs 118.13 µs 118.25 µs]
                        change: [-23.414% -23.315% -23.220%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 2 outliers among 100 measurements (2.00%)
  2 (2.00%) low mild
sqrt_velu/ell=587       time:   [181.18 µs 181.40 µs 181.63 µs]
                        change: [-28.615% -28.509% -28.412%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 4 outliers among 100 measurements (4.00%)
  2 (2.00%) low mild
  2 (2.00%) high mild

After a week of optimising polynomial arithmetic, the biggest speed up I can get is simply not doing any polynomial arithmetic.