What I can indeed tell you about is how Si operates in conjunction with other cognitive faculties of the Ne-Ti temperament. Most naturally it is used as a cognitive storage for impressions and instinctual tendencies that are typically subordinated to the more influential cognitive functions. For instance, Si is highly likely to create distorted impressions that conform to the notions propounded by Ne and Ti. For example, for this type, it is quite natural to pay close attention to what is easily perceived by imagination (Ne) and in what structure is easily perceived (Ti). Thus, when an ENTP is attracted to a certain notion favored by Ne and Ti, Si often concocts impressions and distorted perceptions that support the notion that the person of this type is attracted to.
Inferior Si prompts a natural tendency that is altogether different from that of inferior Se. It often aids a person in creating perceptions that are desired by his or her unconscious rather than what was inspired by an observation. For that reason, one may argue, the inferior sensing of the ENP type is likely to be more prone to distort impressions than the inferior sensing of an INJ type.
Please take a careful note of the following; the above description merely describes a natural tendency and not a behavior that all ENTPs must engage in. In other words, this is a cognitive behavior that simply arises the most naturally in this type and may be superseded by other cognitive tendencies that any ENTP person may have learned as a result of nurture or a conscious effort to act in a way that's different from his natural manner of action.
I suspect that I did not give you the kind of an answer that you were hoping to receive as I do not see typology as a study of behaviors and any behavioristic conception of this discipline engenders irresolvable contradictions. I've commented on the matter in greater detail here.
Cognitive tendencies is the best we have to work with. Although it is true that the way a person thinks often greatly influences how he behaves, however, many other influences also have a say in what he does. Thus, our chances of drawing an accurate correspondence between a certain cognitive faculty and a particular behavior 70, 80 or 90 percent of the time are very slim. For that reason, I recommend that we refrain from doing so altogether.
In other words, the behavior you describe can be induced by a poorly controlled inferior Si, yet I would not go ahead and say that such behaviors constitute the very definition of this cognitive function.
In the end of the day, her conception of temperament consists of cognitive tendencies more so than behaviors and this is what safeguards Thomson's work from the Keirseyan fallacy.
When I was discussing behaviors in my profiles, I had two assumptions in mind.
1. The behaviors are displayed by a person who has the freedom to be true to his temperament.
2. These behaviors are displayed within the context of Western culture.