The consistency condition demands that new hypotheses agree with old ones, leads to the preservation of theories based on seniority, as opposed to quality. Counterinductive theories reveal information that couldn't be obtained otherwise. Thus, the generation of new theories is beneficial, and the preservation of uniformity is detrimental to science at the macro and individual level.
From the outset, the issue of consistency can be disproven rather quickly, when one considers the contradictions involved in reconciling Keplerian and Galilean laws (free fall) and Newtonian ones.
The inconsistency mentioned here is logical, not necessarily experimental.
Additionally, the inconsistency does not necessarily have to exist in the relation between two theories, but can also exist between the consequences of the theories.
T′ is consistent with a finite number of these phenomena (F), inside of some margin of error M.
Consider a theory T′, that accurately explains the phenomena of domain D′:
T′ is consistent with a finite number of these phenomena (F), inside of some margin of error M.
Any alternative theory that contradicts T′, outside of F (the accurately described phenomena of D′), but within the margin of error, is acceptable, if T′ was.
The consistency condition will not allow this, even though that disallowance requires the removal of a valid theory.
One of the critical consequences of this is that the value of a theory is placed in the aspect of it that is unconfirmed: General Relativity was unproved for a while after it was presented, and if the contradictions it presented were what it was judged on, it might have died.
The proof of the insufficiency of this method is shown by the fact that, had one swapped out the older theory with the younger one, the younger one would be protected by the consistency condition.