What do you mean by "wrong"?
When transcribing music for the piano, the goal is to be faithful to the original while ensuring that the transcription is playable and "pianistic", which is a term-of-art that is not very easily definable.
It sounds like you're asking about the rules of traditional Bach-style four-part-harmony. (Or really n-part harmony, where 1 < n < 7, but "four" is the traditional number, so I'll use that throughout.) But it's not "wrong" to write music that breaks these rules--the rules themselves aren't any sort of all-encompassing "Laws of Music", or even really "rules"; they are simply reverse-engineered principles that reflect the implicit voice-leading conventions generally used in Baroque-era (and, to some extent, Classical-era and pre-Baroque) music. These rules are "broken" even in some Baroque music; in other styles and eras, particularly the early Modern era, traditional four-part harmony is often so far from the composer's intentions that it's absurd to look for "violations" of the rules, since the rules simply don't apply. For instance, the 12-tone music of Schoenberg followed a completely different set of "rules" that had very little if anything to do with traditional voice-leading.
So the fundamental question when writing a transcription is, what style is the original music in? In your case, it's somewhat unlikely that the solo guitar piece is a work of Baroque four-part harmony. So it's unclear that you should try to make your transcription follow those rules.
Transcribing an actual Baroque-style piece using four-part-harmony
If the guitar piece is written in the Baroque style, then you may be able to avoid voice-leading "errors" by transcribing the notes exactly in their original registers. This may not produce a very "pianistic" result, however (as noted in other answers).
It's possible that the guitar piece stylistically emulates Baroque four-part harmony without actually following the rules to the letter. This can happen due to the limitations of the instrument itself (certain voicings may be uncomfortable or impossible on a guitar). In this case, it may be a good idea to "fix" the music as you're transcribing it.
Here's what must be preserved from the original, if it's in four-part harmony:
- The bass line
- The top melodic line(s)
- The independence of voices that are clearly intended to be heard independently in the original
Do you really understand the four-part harmony rules?
Of course, you may wish to write a piece yourself in the Baroque style, or you may wish to make a "Baroque-ified" version of a guitar piece, perhaps as part of a cutesy series of variations on a theme. In this case, you'll want to follow the rules as closely as you can, of course.
However, it's not clear from your question that you really understand the rules. What does "doubling the root" have to do with "parallel octaves"?
Let's take a step back. What do we mean by "parallel"? If we simply mean playing octaves as sequential chords on the piano, then that is not merely permissible (in nearly any style) but quite common. The four-part-harmony meaning of "parallel," however, has to do with parallel voices. In fact, all the rules of four-part-harmony are concerned with maintaining the independence of separate voices. The reason parallel fifths and octaves are discouraged is because it makes two voices momentarily blend into one voice, making it harder for the ear to distinguish them. (If you don't know what "voices" are, then you really need to spend more time studying Baroque music and theory before trying to write four-part harmony, but the short version is that they're the independent melodic lines collectively forming both the harmonic and the polyphonic structure of the piece.)
So on a keyboard instrument, "parallel" octaves can either meaning playing a single voice as an octave for emphasis (i.e., the octaves really "belong" to only one "voice", conceptually), or inadvertently merging two separate voices into one single voice by writing them in parallel an octave apart for two or more notes in a row. The first case is of course fine, because there is no loss of independence between voices. The second case is where the "problem" arises (which of course is only a problem when attempting to emulate the Baroque style).
Similarly, when you speak of "doubling" the root, playing the bass voice as an octave does not count as "doubling the root." In order to double the root, you must have two separate voices land on the same note (typically in different registers).
Now, as noted in another answer, parallel octaves only arise in sequential chords; that is, an octave on its own is never considered "parallel." This is because the independence of the separate voices is maintained by their independent motion. Two voices can easily move through the same note (in different octaves) without moving in parallel. Note that if this were not the case, four-part harmony would be impossible: triads only have three notes, so most chords require something to be doubled!
How to break the rules by doubling the root
So here are the necessary and sufficient conditions for you to be "breaking the rules" by "doubling the root", as per your question:
- The original guitar piece must be using (or emulating) n-part-harmony. Alternatively, you may make a conscious decision to follow these rules even if the original piece does not.
- The root must be doubled in separate voices (e.g. the bass and soprano lines, i.e. the top and bottom voices).
- The same voices must play the root in two consecutive chords (e.g. moving from an F-major chord to a G-major chord, the bass and the soprano voice both start on F and move to G).