To answer the question of the title, yes, bilingualism is beneficial in theory.
However, in practice, it is not so. In America we do not have a true bilingual system of education, so it is difficult to judge the success of bilingualism based on our public foreign language education. With about 90% of American High Schools offering and sometimes requiring a foreign language, the benefits of bilingualism (if immersion and high proficiency was not required) are observed to be limited at best.
Moreover, fully bilingual societies still have the same fluid intelligence overall. The nigh-universal Flynn effect isn't improved as societies become multilingual.
Moreover, if you consider my reasoning biased by social stratification, let's examine a more equal set of multilingual groups. Consider (a) the computer science major in college where each student is tasked with learning and using in detail several computer languages of many differing types and forms and (b) the foreign language major. They score well on IQ tests, but so do many other majors not intrinsically requiring immersion into multilingualism.
Yes, switching the languages of thought is a non-task. I do not think (but I cannot prove) there are any advantages to the non-task.
In Computer Science a diverse set of languages from multiple paradigms are taught as a general rule of thumb. This is done so that one is familiar with all the basic concepts of linguistics (declarative, imperative and functional come to mind). In spoken languages anthropology would point you to the most diverse set of languages to build a multilingual set of knowledge.
Though this is anecdotal, my IQ as measured by a trained psychologist (both times) did increase a small amount (5 points) after my training in Computer Science was finished.