It never ceases to amaze me how addicted the media is to the so-called "Sapir-Whorf hypothesis." Follow the link, and you discover that there is actually fairly narrow evidence for this claim (at least the strong version of it). Look into actual experiments on linguistic relativity, and it becomes even weaker.
That phrase "often considered" makes it sound like it's a majority view among linguists, when it's actually a quite fraught contention. It would be like saying, "panpsychism is often considered an explanation of the relationship between mind and matter." Some hold that view, but "often considered" makes it sound more like a widely held position, or even a consensus position. Which it certainly is not.
Some languages have extremely complex ways of indicating tense (making a grammatical distinction between, for example, past events that occurred and then ceased, versus past events that are still ongoing). Some have retained complex grammars for expressing wishes (the optative mood, for example). Other languages have far fewer grammatical ways of expressing tense and mood. Tenses in Mandarin Chinese are often "optional" in a sentence, for example, and in general the mood system of English is a lot simpler than it is in, say, ancient Greek. Ancient Hebrew, strictly speaking, does not have past, present, and future tenses at all (but instead combines context with perfective and imperfective aspects).
Now here's the possibly dangerous leap. "The Chinese think differently about time" (or are less concerned about it). "The ancient Hebrews were very focused on the present and rather feckless about the future." "German and Greek are the only languages worthy of philosophy" (a direct quote from Heidegger). "There are certain emotions in Hopi that cannot be expressed in English." "People speaking a creole are less capable of nuanced rational thought." You can probably see where this is going . . .
Another is grammatical honorifics, which use suffixes to indicate your relationship with the speaker and sometimes also with who, or even what you are talking about.
Habitual readers of fiction tend to be a chore to deal with, as they overestimate their ability to read and understand the thinking of others, inadequatelly communicate as they assume you should be able to read them, and blame any of the inevitable errors on intentional deception from your side.
Which leads me to suspect that such languages must have been evolving, breaking apart and dying all the time.
Those Sami languages are endangered not so much because of their absolute number of speakers as because you can't get anywhere by speaking them. Languages like that tend to die out in favor of more useful ones.
If all the Finns were wiped out and nobody could go into Finland, the Sami would thrive and so would their languages.
Nobody prevents the Sami from reproducing in bigger numbers and living all over Finland (and they consider themselves Finnish citizen, mostly). Also, the usefulness argument contradicts with these languages surviving and Latin having dyed out despite being widely understood (to this day).
The Sami languages haven't "survived" to any greater degree than Latin has. Quite the contrary.
Are you sure you know what Romance languages are? This is not a statement that anyone sane would be willing to support. Why do you think the Romance languages are so similar to each other?
It is conceptually incoherent to say that the modern languages of Sami tribes have survived while Latin hasn't. The Latin of 2300 years ago is not mutually intelligible with the Latin spoken anywhere today. The Sami of 2300 years ago is also not mutually intelligible with any Sami language spoken today. It was even more different, because Latin has a robust written tradition and writing slows language change.
If one has survived, so has the other, and Latin is massively more successful.
On the other hand, if you want to take the position that Latin died out, then the concept of an "endangered language" is meaningless, as we know that all languages will inevitably die in the same way.
The claim that none of these languages are anything like Latin is overblown.