I would never advocate for people to create English language content because of some lingua franca reason. I would say the answer is “it depends” — what is the use, and context of the OER? Isn’t that a design, a pedagogy decision?
It also suggests that the learning and the process is all encapsulated in the OER, leaving out the teacher, the student, working together, the activities around the OER. Maybe the OER is in French and the class activities in Hindi.
And there is also a question of how critical it is for exactness in language, meaning precise translation? Or can the use of automated translations with discussion and interpretation suffice? I am thinking of the (still unblogged) experience I have been part of the last 2 years with the University of Guadalajara. The institution wanted instruction in English, as I understood they have a desire to become more of an international university. Some of the facilitators had partial understanding of Spanish, to me, who has the Spanish skills.. como un niño de tres años. So we managed to conduct workshops in technology and pedagogy in some kind of in between language space, some with human translators, a lot of google translators, patience. Ideally we would be all bi, tri, quad lingual. In practice we are on some mixed spectrum, so we just do the best we can.
I am not sure an absolute language choice is best. Your students may live lives entirely in Arabic spoken culture, or they may be in English (or German or whatever) but it seems like it will likely be some mix. I worry about the lost of culture starting with a diminishing of language.
I would like vote for what ever it takes to make the most interesting and culturally useful OER,.
Of course that’s easy for me to say, more of the world is communicated in my language.