Here is Garrett’s conclusions paragraph giving a descriptive overview of what he has accomplished:

“This paper has progressed in small steps to construct a complete picture of gravity and the standard model from the bottom up using basic elements with as few mathematical abstractions as possible. It began and ended with the description of a Clifford algebra as a graded Lie algebra, which became the fiber over a four dimensional base manifold. The connection and curvature of this bundle, along with an appropriately restricted BF action, provided a complete description of General Relativity in terms of Lie algebra valued differential forms, without use of a metric. This “trick” is equivalent to the MacDowell-Mansouri method of getting GR from an so(5) valued connection. Hamiltonian dynamics were discussed, providing a possible connecting point with the canonical approach to quantum gravity. Further tools and mathematical elements were described just before they were needed. The matrix representation of Clifford algebras was developed, as well as how spinor fields fit in with these representations. The relevant BRST method produced spinor fields with gauge operators acting on the left and right. These pieces all came together, forming a complete picture of gravity and the standard model as a single BRST extended connection. If this final picture seems very simple, it has succeeded. As a coherent picture, this work does have weaknesses. Everything takes place purely on the level of “classical” fields – but with an eye towards their use in a QFT via the methods of quantum gravity, which must be applied in a truly complete model. The BRST approach to deriving fermions from gauge symmetries, although a straightforward application of standard techniques, may be hard to swallow. If this method is unpalatable, it is perfectly acceptable to begin instead with the picture of a fundamental fermionic field as a Clifford element with gauge fields acting from the left and right in an appropriate action. The model conjectured at the very end, based on the related u(4) GUT, is yet untested and should be treated with great skepticism until further investigated. In a somewhat ironic twist, after arguing in the beginning for the more natural description of the MM bivector so(5) model in terms of mixed grade Cl1,3 vectors and bivectors, this conjectured model is composed purely of bivector gauge fields. Although the model stands on its own as a straightforward Cl8 fiber bundle construction over four dimensional base, there are many other compatible geometric descriptions. One alternative is to interpret ⇁ ̃A as the connection for a Cartan geometry with Lie group G – with a Lie subgroup, H, formed by the generators of elements other than ⇁e, and the spacetime “base” formed by G/H. Another particularly appealing interpretation is the Kaluza-Klein construction, with four compact dimensions implied by the Higgs vector, φ = −φ ψΓ ψ, and a corresponding translation of the components of ⇁ ̃A into parts of a vielbein including this higher dimensional space. The model may also be extended to encompass more traditional unification schemes, such as using a ten dimensional Clifford algebra in a so(10) GUT. All of these geometric ideas should be developed further in the context of the model described here, as they may provide valuable insights. In conclusion, and in defense of its existence, this work has concentrated on producing a clear and coherent unified picture rather than introducing novel ideas in particular areas. The answer to the question of what here is really “new” is: “as little as possible.” Rather, several standard and non-standard pieces have been brought together to form a unified whole describing the conventional standard model and gravity as simply as possible.”

see pre-print on http://arxiv.org/abs/0711.0770