Why target enhancement may be doing most of the work
The intuitive picture is that search succeeds because attention blocks out the wrong things. The experiments testing this directly point elsewhere. In one study, singleton distractors appeared in unpredictable colours, so participants could not form a feature-specific suppression template for them — yet evidence of suppression was unchanged [2]. In a second experiment the singleton colour stayed fixed while the target colour varied; without a constant target colour, markers of singleton suppression were effectively abolished [2].
The authors' reading is that the suppression effects observed are largely driven by global enhancement of the expected target colour, possibly combined with a second-order suppression mechanism that does not depend on specific distractor features [2]. This is a proposed explanation rather than a settled mechanism, but it reframes the question: what looks like successful distractor suppression may often be the visible side effect of knowing and boosting what you are looking for.
Distractors still cost you — just not in the way the question assumes
None of this means distractors are harmless. In a 3D search task, search durations were markedly longer when three distractors were displayed, and the effect was most pronounced when a single target and three distractors sat in the same depth plane — it persisted even when the target's position was validly cued [3]. The depth relationship between targets and distractors mattered more than the absolute distance between them, and the authors suggest that within an attended depth plane, irrelevant information produces strong interference [3].
Distractors can also work in the opposite direction. A salient distractor that visually contrasts with its surroundings has been reported to speed search without reducing accuracy, even though distractor objects in general increase reaction time and reduce accuracy compared with fewer or no distractors [4]. So the cost of a distractor depends on the situation, not on a single suppression failure.
Suppression may not happen before attention arrives
Part of why failing to suppress may not be the dominant cost is that suppression may not be as early or as proactive as often assumed. When researchers recorded brain activity while people searched, salient distractors triggered negative event-related potentials (N1pc/N2pc), boosted the steady-state visual evoked potential relative to non-salient filler items, and suppressed alpha-band amplitude on the opposite side of the head — three electrical signatures associated with attention being allocated — even though those distractors did not interfere with behavioural responses to the targets [1].
In plain terms, the brain was still registering and attending to the distractors, yet people's responses were unaffected. The same recordings indicated that stimulus-driven and goal-driven attention occurred together, with the goal-driven effect enhancing and prolonging the stimulus-driven one [1]. That is consistent with the idea that what protects performance is the strength of the target-directed signal, not a clean pre-emptive block on the distractor.
What this means for a cluttered desk or a word-search puzzle
The practical implication is that knowing exactly what you are looking for — a specific colour, a specific word shape — may matter more than trying to ignore what you are not looking for. That fits the finding that a fixed target colour preserved suppression markers while a varying target colour erased them [2].
Context still shapes the cost. In a word-search task, search time was significantly delayed when the search direction was anything other than left-to-right, and when an audio distractor was present on a digital medium; average search time was about 100 seconds for left-to-right search versus about 291 seconds for random search [5]. Those are large differences in a task where the target set was fixed, which is a reminder that layout and sensory distraction can dominate even when target knowledge is constant. The comparison between failing to suppress and failing to enhance has not been tested in that kind of reading task, so how far the target-enhancement account extends beyond colour-based search remains open.
Sources used in this answer
Dynamics of attentional allocation to targets and distractors during visual search
Salient distractors elicited attention-related neural signatures (N1pc/N2pc, enhanced SSVEP, contralateral alpha suppression) without interfering with behavioural responses to targets, and goal-driven attention enhanced and prolonged the stimulus-driven effect.
Testing the contributions of proactive target and distractor templates in visual search performance
Suppression markers persisted when distractor colour was unpredictable but were effectively abolished when target colour varied, suggesting suppression effects are largely driven by global enhancement of the expected target colour, possibly with a second-order distractor-independent mechanism.
Visual search in virtual 3D space: the relation of multiple targets and distractors
In 3D search, three distractors markedly lengthened search durations, especially when a single target and three distractors shared a depth plane, and this persisted even with a valid spatial cue.
Salient distractors facilitate visual search
Distractor objects generally increase reaction time and reduce accuracy, yet a salient distractor can counter-intuitively expedite search without reducing accuracy.
Role of reading medium and audio distractors on visual search
Visual search time was significantly delayed for search directions other than left-to-right and in the presence of an audio distractor on a digital medium, with average times of about 100 seconds for left-to-right versus about 291 seconds for random search.