Take Minecraft for a second. It has been lots of things for different kids (old kids too). YouTube is overflowing with HowTo videos on every aspect of MC.
There are different types of kids (and viewers of those videos). High-end performers that would have discovered most of those things on their own and who spend time trying to out-do the videos, becoming inspired to try more things on their own. Low-end copycats that don't have any imagination of their own, who are the same to follow Lego instructions to build the Harry Potter castle as it was blindly put before them in an instruction book. Page by page, part by part. Then there is the usual large bunch in the middle that span and vary the two extremes.
Now for the low-end kid (as I have evilly labeled them), do they get any value or utility from doing Legos by-the-book, or only doing in MC what they just saw in a video?
A big part of me says no. I am one that really misses and believes in the value of going out in the garage or backyard and just tryin' stuff. So many youngsters want the final working product and no longer have interest in figuring it out on their own. I invest in FLL for some hope to get kids hands-on something and something with no set answer key in the back of the book.
Yes, there is some utility in "keeping kids involved" or "getting them hooked and interested", that is indeed better than passively watching TV and their childhood waltz by. If one kid gets inspired by EVLessons or the others, I could never campaign against it.
So if a team or kid took on the Mindstorms platform and each FLL season on their own, and stumbled and scratched their way to better-than-average - then I think they will have learned and been convinced of a way of thinking, that propels them way forward towards excellence in their next 10-15 years.
If a team or kid opens up the Mindstorms box and immediately rushes out to copy what is available from the internet (or is spoon fed from a well-meaning but over-involved mentor), and they too manage their way to a better-than-average result - then I think they have a decently likelihood of always being just better-than-average.
Now there is one other angle. Let's say you knew you wanted to be a programmer. Is it really needed these days to learn assembler and twos-compliment storage methods, etc? To learn "from the ground up" so you "appreciate" higher-level languages and how not to write code that is wasteful of memory or resources, etc. The journeymen, apprentice, master concepts.
If we always state that every master has to start from square one - then we stifle overall innovation. Only by building-on-the-shoulders of other masters that already conquered assembly and compilation concepts - do we rapidly progress.
But. Here is where I think FLL has a different role. FLL (and the age group it targets) is NOT already slotted to become programmers or robot designers, or whatever. It is a great EXERCISE to start to developing muscles and skillsets that one needs to design and problem solve. And if we want them to be "masters" of problem solving (no matter what field and future they behold), then I think you get the best bang for the buck by starting them at square one.
Only if square one makes them turn towards square zero, and these barriers to "getting going" can be mitigated by the likes of EV3Lessons or Lego tutorials or even Mr Dude35 - well then, that is certainly OK and better overall.