Don't over-constrain search space.

Pardon, but the developers rapidly came to appreciate the theory behind the sieving polynomial itself, this morning. The short explanation is that the immediately-previous release, while working, is still over-constrained in its potential solutions, but this release fixes the issue.

Full Changelog: v5.2.2...v5.2.3

sha1sum results:
bab6482fe9fb7a105b026e00558535bb748fc85e FindAFactor-5.2.3-cp310-cp310-manylinux_2_35_x86_64.whl
1877ca5715feec02fe2e372bb5ff9c886552ae62 FindAFactor-5.2.3-cp312-cp312-manylinux_2_39_x86_64.whl
6b31317fadbfb5fbf64cc92de5172d303f852ce7 FindAFactor-5.2.3-cp312-cp312-win_amd64.whl
7e56305882963187cd5707e88584c10827016cda FindAFactor-5.2.3-cp313-cp313-macosx_13_0_x86_64.whl
b67d3536993587a7cf1bd363538593d947bcbff7 FindAFactor-5.2.3-cp313-cp313-macosx_14_0_arm64.whl
0eb126cd8fc8223c7f6891b94c75fe8f20d5a85e FindAFactor-5.2.3-cp313-cp313-macosx_15_0_arm64.whl
c8817d4798565df6435340600d8fb94374f0609d FindAFactor-5.2.3-cp38-cp38-manylinux_2_31_x86_64.whl
9dd89b4ce46855c854f879797dd24f80efbed7e8 findafactor-5.2.3.tar.gz