The technical writing necessary for science must have only a single exclusive goal: transmittance of information to the widest audience possible (for widest possible validation, experimentation and for other reasons reported in the article under advantages).
The listed disadvantage are in fact only due to the fact that there are still some who cannot/would not publish in the language that has became a standard (like the fact the Chinese did not report their first results in English).
It a great advantage, facilitating untold benefits to humanity that one language became the standard (at least for science). Having the contrary solution where journals in language, or even each dialect would require translator, hence, mistakes, would mean simply costs: tangible as well as intangible which do not benefit anyone in the long run.
Of cause, for cultural wealth, modes of thinking/psychology and other many valid reasons it is desirable to know as many languages as one can, but there should be one language as a pure communication medium among people, for science and at best for non-science as well. Our brains have no problem to master 5-12 languages at a level that one can read literature and personally enjoy benefits of such knowledge.
