The problem that arises at the micro-level is the hysteresis phenomenon commonly associated with how a magnetic flux (B) is induced by a magnetizing force (H), illustrated on the Encyclopædia Britannica Web site as follows:
As the upward arrow illustrates, when a magnetizing force is first applied, there is almost no change in the induced flux. It is only after the continued application of that force has accumulated to a certain level that the flux begins to change, after which it rises relatively quickly in response to additional force. In a similar way a violinist needs to accumulate some initial application of force before the string can vibrate at all, and it is only after that vibrating state is achieved that it makes sense to say that the expenditure of energy is explicitly under the control of the performer. This is particularly problematic if that initial energy level needs to be low (producing a sound of low amplitude), because that energy level may be too low to set the string vibrating at all. (For example, this is what makes the very opening of Richard Wagner's Lohengrin so problematic for both the conductor and the string section being conducted.) My point is that, before you can worry about the strategy for controlling the energy of your sound, you have to satisfy the precondition of having the sound at all, after which you need to quickly align the sound you have to the energy level of what you want.
That process of alignment is probably driven by some sense of expectation coming either strictly from the soloist or through some agreement between soloist and conductor. In other words there is this delicate matter of having a clear sense of what you want to hear, which must be reconciled with the (likely) possibility that the very first thing you hear (because of hysteresis) is not what you want and must therefore be corrected almost immediately. Furthermore, in the case of a concerto, that sound you want must eventually become part of a larger whole, integrating with the sound of the ensemble while maintaining its own identity. Thus, at the macro-level performance is very much a matter of ongoing alignment that involves not only the sound of your own instrument but the sound of the entire ensemble including your instrument.
This is a rather challenging set of actions to strategize, which is why practice and rehearsal are so important. To some extent your capacity to know what you want begins and the micro-level and gradually works its way up to the macro-level as you become more familiar with the "text" through practice and rehearsal. Here, too, there is probably a hysteresis effect. When the rehearsal process begins, progress is slow; and the "returns" are relatively small. However, eventually (sooner, rather than later, in most professional settings) the rate of progress improves, probably because the micro-level strategies that are applied in one setting turn out to be applicable throughout the entire composition. Put another way, after a certain amount of rehearsal one acquires a "repertoire of strategies," after which strategy escalates to a "meta-level." The good news is that most of the repertoire (like the Bruch concerto) is sufficiently familiar that one is already prepared with ideas about strategies even before the first efforts at performance; and, even when the music is not familiar, one will tend to draw upon personal performing experience to propose "trial strategies." If not all of those trials go according to plan, one can always compensate, which is usually far easier to achieve in the immediacy of rehearsal than the more daunting task of coming up with a new strategy from scratch.
Needless to say, this is all highly hypothetical. To some extent these very ideas are the product of micro-strategies trying to grow into macro-strategies. Such growth, however, is just another way to look at the results of rehearsal, which is why the ideas have been presented in this particular "Rehearsal Studio!"