The short, general answer
All things (all voices) being equal, solution #1 is the correct technical approach. The shorter note should interrupt the longer one, but then continue to be held (allowing for additional interruptions) for the duration of the longest note.1,2 Solution #2 is appropriate in some situations (see below). Solution #3 is never appropriate.
The short, specific answer
For the measures given, I would slightly emphasize the long notes to help sustain them. Since they are being held (i.e., are static), the moving voice will naturally be perceived as superseding them in prominence. The allows the moving voice to be played dynamically just below the emphasis level of the long note. In that way, the re-articulation, slightly de-emphasized, will be heard as part of its own voice, but its sustain will be heard as the sustain/decay of the original, emphasized articulation.
Further discussion
The important distinction in this question from other similar ones is the focus on maintaining voice independence across the pitch repetition.3
It is customary when playing polyphonic music on the piano that voice independence is maintained by variations in dynamic level and articulation (touch) between voices.4 In the following discussion
*first voice* = the polyphonic voice first encountering the repeated pitch.
*second voice* = the polyphonic voice responsible for rearticulating the repeated pitch.
Voices distinguished by dynamics
Louder voice = first voice
It's a bit easier if the first voice is louder than the second, because the second voice's sustain more convincingly sounds like the natural decay of the first voice.
Louder voice = second voice
Should the second voice be louder, then a case can be made for releasing the note in the second voice rather than trying to preserve the sustain of the first voice: the interrupting, louder voice taking precedence.5 This is option 2 in the OP.
Voices distinguished by legato/staccato touch
Legato voice = second voice
When the second voice is the legato one, it allows for a blending of the natural decays of both voices.
Staccato voice = second voice
As with the dynamics case, if a staccato voice is interrupting a legato one, the a case can be made to curtail the first voice (but see note 4).
1 This is also true if a long note is interrupted by another long note. The interrupting note is held through all subsequent interruptions, even past the length of the original note, until the longest "current" value is exhausted.
2 The same principle is discussed in other similar questions, such as overlapping note on same staff.
3 For example, in the question mentioned in note 2, it is important to maintain the distinction between the held melody note and the accompanying chords that rearticulate it. If the chords are too loud, the effect will be as though the melody comprises repetitions of the same note.
4 In Bach's two-part Inventions, for example, students are often introduced to this concept by being taught to play sixteenth notes legato and/or louder while accompanying eighth notes are played staccato and/or softer.
5 In some cases, even when the second voice is louder, the decay of the first voice can be maintained. This involves making the second articulation and then briefly allowing the damper to just barely contact the string -- damping it somewhat, but not fully. This can be achieved either with the pedal, by a half-release, or by half-releasing the key itself, but does require technical circumstances that allow for such careful control.
<< { g8 e r a } \\ c2 >>
). I agree with the answers that option (1) is better, but I’ve heard this alternative, too.