No, that’s ridiculous.  It would imply that the frequency analysis is wildly incorrect.

However, what do we get as a result?

These things are typically used with a keyword, it makes reversing the process easier.  One of the ways to solve K3 is with a keyed-columnar transposition in part of the deciphering method.  The key can be numeric or a word and these can be ordered and reordered accordingly.

I used KRYPTOS as my keyword and:

OBKR
UOXOGHULBSOLIFBBWFLRVQQPRNGKSSO
TWTQSJQSSEKZZWATJKLUDIAWINFBNYP
VTTMZFPKWGDKZXTJCDIGKUHUAUEKCAR

became

OSDR
MOIBSINZLAKEGGFIWRKKKPFIUZUSKBN
OZHSWBFXWUOGWBOAATDFNGTUWKLYHJE
TZRPUKKQXVVLLCSTQTBUAJJQTSDRQCP

Oh well.  It didn’t exactly break my heart when this didn’t work.  It’s not ridiculous to think that transposition plays a part in K4 but an expectation that reversing a transposition alone will reveal plaintext is misguided at best.  Something was done to the text of K4 and whether it was a masking technique oranother ciphering method, it will require solving at least two cipher methods to get to the original K4 message.

–Kryptosfan