One of the keys to understanding the New Testament (NT) use of the Old Testament (OT) may be the recognition that when a NT author draws upon an idea found in a particular OT passage, it does not have to be the main idea of that passage to be usable. The contemporary assumption (often not articulated) that it has to be the main idea of an OT text to be legitimate seems to be a key stumbling block for people studying the NT use of the OT. The tendency for people to focus only on the main idea of a text (rather than also upon sub-themes) may also explain my present discomfort with the sense / referent distinction made by various authors. The sense / referent distinction seems to assume a single sense for a verse that is akin to an exegetical idea of that verse.
But sub-universes of meaning do exist in every text, as Berger and Luckman observe. Sometimes the NT authors draw upon these sub-meanings rather than upon the main idea that an OT author was intending when penning a particular passage.
This insight also explains the times when it seems as though atomism is taking place in the appropriation of an OT text by a NT author. It could be simply that a NT author is drawing upon a sub-meaning in a text he is employing, which is acceptable (in my opinion) because it is part of a broader theme (e.g., Matt. 22:32, Exod. 3:6,15-16). Ironically, this allows for a thin use and a thick use to be taking place simultaneously! It is thin inasmuch as the NT author may be drawing upon a small sub-theme of an OT text, but thick inasmuch as the NT author is knowingly tapping into a fuller thematic pattern that runs throughout the OT and beyond.
Parable studies may have contributed to the problem of assuming that there must be one—and only one—meaning in every text, without attending to possible sub-themes in given texts. The single meaning approach was still the dominant view of how to interpret parables during the time that the sense / referent distinction was beginning to be employed in explaining the NT use of the OT. The teaching and practice of homiletics may also have contributed to this assumption. Many who study the NT use of the OT are also preachers who preach using the big idea approach to preaching. But such preaching assumes that there is one and only one main idea in every text.
But we regularly draw upon sub-meanings when systematizing theology in regards to topics that are not explicitly addressed in particular texts. For example, we talk about how God is a creative God who loves beauty, and reference Bezalel's God-given creativity (see Exod. 31, 35-38), even though God’s creativity and beauty isn't the main idea of any of the texts in which Bezalel is found. But drawing upon this sub-theme—an idea that is actually there—is acceptable because it is part of a broader thematic pattern found in the biblical canon. We also sometimes draw upon sub-meanings when we apply OT commands to today. For example, is there not a sub-theme of modesty in the instruction about care in going up to the altar in Exod 20:26? It is certainly not the main idea of that text, but still illustrates a sub-theme that fits with other teachings in the Bible about modesty.
Conclusion: We must allow that when the NT authors use the OT, they sometimes refer to patterns or themes that appear at some level in particular texts they are citing, even if a particular theme is not the main idea of the particular text employed. The NT authors sometimes draw upon sub-meanings when they use OT texts.
 As one example, see Darrell L. Bock, “Single Meaning, Multiple Contexts and Referents: The New Testament’s Legitimate, Accurate, and Multifaceted Use of the Old,” in Kenneth Berding and Jonathan Lunde, eds., Three Views on the New Testament Use of the Old Testament (Grand Rapids: Zondervan, 2008), 112-115.
 Peter L. Berger and Thomas Luckmann, The Social Construction of Reality: A Treatise in the Sociology of Knowledge (Garden City, NY: Doubleday, 1966), 85-86.
 Pioneered by A. Julicher, Die Gleichnisreden Jesu (2 vols.; Tubingen: J. C. B. Mohr: [Paul Siebeck], 1888-1889).