No, it doesn't contradict anything. Elite schools take the most qualified students with the highest numbers. They are not going to take someone with a 168 over someone with a 173, even though both would be more than capable of doing the work, just because someone has a great PS or WE. We're not talking about a difference that isn't even statistically significant. A 170 and a 172 are in essence the same score, so it's a pretty useless example.
A lower ranked school, which does not mean a school where students have LSATs in the low 150s, but perhaps somewhere like GW, would be much more likely to take a gamble on a student with a 163 over someone with a 168 if they have some sort of compelling non-numerical reason, even though that is the same 5 point spread a T14 school probably wouldn't overlook. The applicant pools at lower ranked schools are less cloistered and more diverse in terms of what type of applicants they are dealing with.
Of course schools say they take all sorts of things into consideration, blah, blah, but if elite schools actually accepted even a quarter of their applicants outside of their numbers (not one point lower or higher), they would not remain the same 14 for long. They realize that, and lower ranked schools have much more freedom to go from 24th place to 26th place and back up again and in turn, to not be so devoted to a certain level of numerical achievement they have to maintain.
Also, those who score in the 150s on the LSAT are not overwhelmingly likely to not make it through law school, or to not achieve success as a lawyer. In fact, those who do not are in the overwhelming minority. Even a majority of Cooley students pass the bar, and we've all heard more than enough about Cooley's weaknesses. It wasn't so long ago that an LSAT in the 150s would get you in at some really highly ranked schools, and LSATs in the mid to high 150s still will in some instances.
This conclusion that a high LSAT equals success or failure as a law student has no evidence to support it. The only correlation between LSAT and academic success is that those who have a higher LSAT SOMETIMES have higher 1L grades. The correlation is not that people with lower LSAT scores fail out of law school.
Your statements are conclusory, and in some cases, inaccurate. In the following case, both!
To bolster your argument that somehow elite schools are blinder to soft factors than lesser schools, your premise is that "Elite schools take the most qualified students with the highest numbers." Is this accurate? I assume, then, that George Bush got into Harvard Business because his 2.75 GPA was mind shattering, or JFK Jr's--what, 155 LSAT--proved he deserved to attend one of the finest law schools in the country. But, we don't even have to use these examples to prove that there is a history of elite colleges of all types weighing things other than statistics in their admissions policies. This information, coupled with the admission on the part of ADCOMS of elite schools that they rely heavily on soft factors to differentiate what are outstanding candidates.
Now, I would like to believe that schools like George Washington are uninterested in their ranking, and would be more likely than Harvard to accept someone with lower numbers, but I don't see much evidence proving this. In fact, evidence seems to suggest that schools without the luxury of high rankings, they are likely to work equally as hard, if not harder, to improve their ranking, and one could not do this without being sensitive to the impact of the statistics of its candidates. Harvard, for example, is less likely to see a fall in rankings than GW because of inertia. After all, even Princeton gets votes in the rankings, despite not even having a law school!
I was amused by your statement that there is no real difference between a 168 and a 170. I happen to agree with you on this point, but this reality further errodes your argument. If there is no difference between the scores, then this only increases the likelihood that soft factors will creep their way into decisionmaking. Now, you may discount its relevance because there is no real difference between the two scores as a reflection of academic ability, but a 2 point difference on the LSAT is a 2 point difference in the statistics which in part determine the rankings. USNWR does not care whether the 2 points were at the top of the scale or at the bottom. Schools know this!
But let us assume that schools never considered personal factors unless the LSAT differential was meaningful, lets say a 15 percentile difference. Then, it is likely that you must give elite schools a pass altogether, since most of their competitive applicants are at the "irrelevant" end of the spectrum. Your statement in this universe has little meaning, for the schools are forbidden from taking soft factors into consideration much less than they simply don't do it by choice. Regardless, we all know schools do take meaningless GPA and LSAT differences into account. Whether this is wise is another argument altogether.
But, of course, GW and Harvard consider far less significant percentile differences than you suggest, and even so, we are talking about Top 25 schools only. It is reasonable to assume, if you are correct that is, that when you get to the bottom of the rankings, schools would routinely deny admission to those who score significantly higher on the LSAT over those who performed worse. We tend not to see this. In fact, most persons, including ADCOMS, view non-elite schools--particularly Fourth Tier schools, as being more numbers driven than the TOP 25 generally, and the Top 5 specifically.