When mixing math fonts with the range option one problem are the fontdimens: currently unicode-math uses the fontdimens of the last font set with \setmathfont and quite often these don't fit as the trick to use as last math font \setmathfont[range=\int]{mainmath font} has side-effects and destroys previous range settings.
When mixing math fonts with the range option one problem are the fontdimens: currently unicode-math uses the fontdimens of the last font set with \setmathfont and quite often these don't fit as the trick to use as last math font
\setmathfont[range=\int]{mainmath font}
has side-effects and destroys previous range settings.Imho it would be good if one could store and restore the fontdimens, perhaps even one-by-one. One first try I made (for xelatex) with this is here: https://tex.stackexchange.com/questions/398159/nth-root-kerning-problem-with-unicode-math-font-override.