Well, after a break, I think it is good to go back to the topic of the Covenant of Works. In my review of a Piper sermon on Rom. 2:6-10, I have touched a little bit on the Covenant of Works, as I am certain that contrary to Piper's assertion, Rom. 2:6-10 is a restatement of the Covenant of Works, and flows much smoother than read that way in its context rather than to read it as believers fulfiling those conditions by works done through faith (a dangerous proposition indeed bordering on Romanist synergistic soteriology).
With this brief introduction, I would like to go deeper into this topic; to indeed show that the Covenant of Works is indeed biblical. In fact, I hope to show that not only it is biblical, denial of it logically undermines the foundation for the doctrine of the free grace of God, and thus showing it to be an important (though non-cardinal) doctrine indeed.
In Robert Reymond's A New Systematic Theology of the Christian Faith, Reymond states the Covenant of Works this way:
... it [the covenant between God and Adam] is called either a covenant of works ([Westminster] Confession of Faith, VII/ii; XIX/i) or a covenant of life ([Westminster] Larger Catechism, Question 20; Shorter Catechism, Question 12), the former chaterization emphasizing that the confirmation in righteousness which God would give Adam upon the latter's successful sustaining of his probationary test He would necessarily give to Adam in justice and that what Adam would receive he would receive as reward or merit for his obedience, the latter characterization specifying the nature of the reward which Adam and his posterity would receive if he obeyed God. (p. 431)
Reymond further defends the concept of the existence of the Covenant of Works in the few pages in this book of his against various attacks. The first thing he does is to show the concept from Scripture, and this he did excellently. First, he showed that the Hebrew word for covenant does not have to be present for a covenant to be present (the Davidic Covenant in 2 Sam. 7 cf Ps. 89:19-37). Secondly, the covenant elements (parties, stipulation, promise, and threat) are present. Thirdly, he defended the traditional rendering of Hosea 6:7 on exegetical grounds to show that it reads that Adam transgressed the covenant. And fourthly, he shows that the New Testament parallels between Adam and Christ (especially in Rom. 5) imply that just as Christ was the federal (foedus: covenant) representative of the New Covenant, so also Adam acted as a federal representative of a covenant arrangement (p. 430).
All of this are solid reasons for the existance of a covenant arrangement between God and Adam before the Fall, and the fourth reason should have the most persuasiveness among Evangelicals who tend to focus more on the New Testament, and especially so since the Apostle Paul stated it so clearly in the book of Romans (chapter 5). Therefore, exegetically, there should be no contest as to the existance of the Covenant of Works between God and Adam.
Reymond uses the works of Daniel P. Fuller as a foil as one which seeks to undermine the existance of the Covenant of Works. Noting that Fuller sees grace in God's dealings with Adam, insisting upon a 'continuum' of divine grace in all God's dealings, Reymond shows that Fuller, for all his talk about grace, actually "winds up with true grace nowhere and a kind of works principle everywhere" (p. 431). This is definitely true, since if grace is everywhere, then the differances between the efficacy of various 'graces' must lie in something else which is not grace (i.e. the instrumentality of faith, and faith as a human autonomous work). Although Reymond concedes that Fuller is quick to point out that "such good works are not meritorius" (p. 431), this just shows that an insistance on grace everywhere logically leads to the denial of the Reformation principle of Sola Fide. Just because Fuller refuses to be consistent with the logical direction of his theologizing does not make such an error less dangerous.
Reymond later touches a bit further on the 'Fuller approach', quoting Meradith G. Kline who showed that the Fuller/Norman Sherpherd theology, consistently developed, would result in the fact that
... the work of obedience performed by Jesus Christ did not merit a verdict of justification from his Father. The justification of the second Adam was not then according to the principle of works in contrast to grace, but rather found its explanation in the operation of a principle involving some sort of grace — a grace required because of the inadequacy of Christ's work to satisfy the claims of justice. (quoted in p. 432)
And this indeed is devastating to the doctrine of justification. Reymond correctly states that "if Christ's obedience has no meritorius value, neither has a penal substitution been made for our sins nor is there a preceptive righteousness available to be inputed to us' (p. 433). It is also interesting to note that Kline has made a similar observation regarding the denial of the Covenant of Works. Building on the parallel between Adam and Christ as the Second Adam, I have combined it with elements of Christology as previously written,
If truly indeed the Covenant of Works does not exist, then Jesus could not himself merit eternal life while living on earth, and thus he is in some sort of salvation limbo as with regards to his humanity. Is anyone going to seriously say that Jesus did not merit salvation with his sinless life; that Jesus because of His bearing a human nature cannot go into heaven unless He Himself died on the Cross to earn salvation for his human nature? Yet, that is what the denial of the active obedience of Christ, and to a lesser extent the Covenant of Works, will lead to. If righteousness can only come to Man via the death of Jesus on the Cross, then Jesus with regards to His humanity must also be redeemed via the Cross, which is a blasphemous notion indeed.
Enough has been said to prove the existance and importance of the Covenant of Works. Now what is the relation between the Covenant of Works and the Covenant of Grace?
The Covenant of Grace is often used to describe the historical expression of the covenant that God the Father made with God the Son to save the elect before the foundation of the world (the Covenant of Redemption), in the entire plan of salvation running throughout Scripture, promising eternal life to all who believe in Jesus Christ by grace through faith. That such a covenant exist is exceedingly plain from Scripture and is automatically included when the topic of Election and Predestination is believed upon. As such, there is no contest, at least within Calvinists, as to the existance of the Covenant of Grace. Less discussed is the relation between the Covenant of Grace and the Covenant of Works. However, as I will show, they are actually logically connected. Without the Covenant of Works, the Covenant of Grace would be less stable in its foundation.
Now, the reason why this is so is along the same line as to how Reymond states why Fuller is in error. Without the Covenant of Works, there is a logical deficiency in the Covenant of Grace. Besides the undermining of the active imputation of Christ' righteousness to our account, as we have shown earlier, a more fundamental problem is that denial of the Covenant of Works undermines the justice of God in His dealings with Man. How can Man be said to fall foul of the justice of God when the Covenant of Works is denied? (Yes, you can do based on Scripture; what I am saying is that it is illogical for these biblical texts to be there if you do deny the Covenant of Works)
Of concern also is with regards to God being the author of evil; in the sense that He is culpable for it. For if the Covenant of Works is denied, then God is not fair towards Man. Yes, God does not need to be fair; if He is fair, we ought to all go to hell in judgment. That however does not solve how God can be said to not be culpable for evil, seeing how is it that no one is given a chance. It can of couse be 'resolved' by semantics (by definition, what God does is good and not evil), by postulating secondary causes, which are all legitimate. However, the problem is not well resolved. If the Covenant of Works is believed upon, it is sincerely believed that a stronger case could be built to show that God is not culpable of evil (author of evil), since God has given everybody a sincere chance, and it is hypothetically achievable but for the fact that we have all sinned in Adam as our covenant representative.
In conclusion, we can see the importance and the Scriptural proof for the Covenant of Works. Although not cardinal, nor even important on the surface, the logical implications of this doctrine has shown it to undergird and support many other important doctrines. As such, let us learn of it and hold to it, even when the New Covenantal Theologians in ignorance try to get rid of it.
 Robert L. Reymond (1998), A New Systematic Theology of the Christian Faith, 2nd Ed., Thomas Nelson Publishers.