Found a minor design inconsistency with 羽 (U+7FBD) ~and possibly 翻 (U+7FFB)~. The last pickup stroke does not touch the outer 𠃌 component in similar JP glyphs with the 羽 component (JIS90 forms shown), and yet for these two characters, the CN glyphs are being used for JP.
However, for 翻 (U+7FFB), in Kozuka Gothic, the right 羽 component has the last strokes touching the 𠃌 parts because it's space constrained; the only difference between it and the CN glyph is the design of the top-left 釆 component. Replacing the CN glyph with an unreleased JP glyph based on KozGo will not solve the design problem of 羽, only that the 釆 component, if the JP glyph was being used for JP and CN, could look a bit "inappropriate" for CN handwriting standards, so that's another matter not relevant to this design issue of the 羽 component.
Found a minor design inconsistency with 羽 (U+7FBD) ~and possibly 翻 (U+7FFB)~. The last pickup stroke does not touch the outer 𠃌 component in similar JP glyphs with the 羽 component (JIS90 forms shown), and yet for these two characters, the CN glyphs are being used for JP.
My suggestion for 羽 (U+7FBD) is to replace the TW glyph with an unreleased JP glyph based on Kozuka Gothic that can be mapped to JP, TW and HK, and keep the CN glyph only for the CN locale.
However, for 翻 (U+7FFB), in Kozuka Gothic, the right 羽 component has the last strokes touching the 𠃌 parts because it's space constrained; the only difference between it and the CN glyph is the design of the top-left 釆 component. Replacing the CN glyph with an unreleased JP glyph based on KozGo will not solve the design problem of 羽, only that the 釆 component, if the JP glyph was being used for JP and CN, could look a bit "inappropriate" for CN handwriting standards, so that's another matter not relevant to this design issue of the 羽 component.