True or not, here's a very practical line of reasoning:

* It's generally accepted that gaining reasonable fluency in one's first acquired natural language takes about four years.

* It's generally accepted that gaining reasonable fluency in each subsequent acquired natural language takes about two years.[0]

* Learning a clean "constructed" language is reasonably efficient. Fluency in Esperanto or Lojban, for example, can be attained in as little as one year.

* Thus to learn, say, Swedish, it's fastest to learn Lojban first in one year, and then Swedish is a subsequent acquired language and only takes an additional two years.

* Conclusion: It's faster to learn "(Esperanto|Lojban) Swedish" than it is to learn Swedish alone.

I don't entirely subscribe to this theory, but it's an interesting speculation.

For me, I learned some constructed languages because I wondered if they would provoke the Sapir-Whorf effect. Anecdata: They did.

[0] The rationale is that the first requires un-learning things you've always taken for granted with regards your cradle language. Learning subsequent languages doesn't require that "un-learning, and so is quicker."