Action maps are the worst means of assessing position on the field, except for all the others. Consider the chalkboard plots of Clint Dempsey and Jordan Morris from Saturday's game, courtesy mlssoccer.com.
Now consider the average position map, from whoscored.com, playing left to right.
Morris, Dempsey, and Aaron Kovar are neatly placed upon one another. This isn't to suggest the forwards shared the same space in the run of play. Morris and Kovar ranged both in front and to the back of Dempsey on both flanks, switching sides as is the natural habit of Sigi Schmid flank attackers (and an almost-necessary feature of attacks imbalanced in possession play and footedness). Herculez Gomez exhibits the typical average position of a right side winger/attacking midfielder. For reference, take a look at the position map of another MLS 4-3-3, courtesy Sporting Kansas City (playing right to left).
The three midfielders in SKC's setup are not always so neatly placed on the center circle, but are typically no further upfield or downfield. The wingers, if they switched regularly, might be expected to take up a similar position to the overlapping Sounders forwards of the previous plot. Dom Dwyer (#14) plays higher up. KC features 3 central midfielders covering a range of skills in creativity and defensive activity, but even the most attack-minded (Benny Feilhaber) can step in to account for ~10% of a team's defensive actions over a significant interval of playing time. All this is not to suggest that there is a "right" way to design a 4-3-3, but deficiencies in one part of a formation must be accounted for elsewhere on the field. Feilhaber, for example, puts in about 3.3 defensive actions per game to 1.9 from Andreas Ivanschitz. These aren't huge numbers by themselves, but they reflect a proportionate degree of defensive engagement that goes unmeasured.
The 2016 Sounders have engaged in a prolonged experiment to determine whether, essentially, (1) Dempsey or Ivanschitz in the central attacking midfield can defend adequately alongside Osvaldo Alonso and Cristian Roldan, (2) whether the defensive line can absorb additional pressure, (3) whether the forward band can pick up the slack, or (4) whether a more defensive posture of the withdrawn CMs and fullbacks can leave the offense intact. None of these questions have produced a satisfactory answer.
Option #3 was evident in the first half against Columbus, with credible defensive energy from Morris and Kovar (defining who, exactly, was the final midfielder among Seattle's four attackers in the first 45' would be a useless gesture, no matter the tactical data I have obtained).
In the second half, a high-energy defensive performance from Alonso and Roldan disrupted the Columbus attack in the middle third. Erik Friberg only improved this, upon replacing Kovar.
Note, for example, that Joevin Jones was responsible for ~18% of Seattle's defensive actions (excluding recoveries) in the first half. He did not register a single defensive action in the second half of the game. That, of course, is not a sustainable feature of the defense, but the win against Columbus should nevertheless have the Sounders asking whether their faith in the 4-3-3 demands a Friberg/Alonso/Roldan midfield.
Raw data for this work was gathered from OPTA via whoscored.com.