I replied that for Japanese at least, probably due to symbol input being too tedious. However I think it's worth mentioning a potential mitigation, and maybe even an advantage.
As a mitigation, full-width symbols could be used instead, which are typically the default in Japanese input. Japanese itself is also fixed-width so if done across the board the language itself becomes fixed-width, not just by virtue of a font selection.
As an advantage, some logical symbols, greek letters, other rare characters are easy to input in Japanese mode, and this could lend itself to a more symbol-heavy language design. I already take advantage of this myself with Δ, φ, and τ use selectively in a few projects. Symbols with easy entry may differ by OS, but here are a few other examples that could be useful:
≠, ≡, ∞, ∴, λ, θ, α, β, ・, °, ※