Excerpt  from a few emails I recently sent in regards to milk, dairy fat, and pediatric obesity.  Correct me if I’m wrong, but my primary contention is that we don’t have good evidence behind recommending low-fat milk in pediatric populations.
Email 1
Berkey, C. S., Rockett, H. R. H., Willett, W. C., & Colditz, G. A. (2005). Milk, dairy fat, dietary calcium, and weight gain: a longitudinal study of adolescents. Archives of pediatrics & adolescent medicine, 159(6), 543–550. doi:10.1001/archpedi.159.6.543
Centers for Disease Control and Prevention (CDC). (2010). Effects of switching from whole to low-fat/fat-free milk in public schools — New York city, 2004-2009 MMWR. Morbidity and mortality weekly report, 59(3), 70–73.
O’Connor, T. M., Yang, S.-J., & Nicklas, T. A. (2006). Beverage intake among preschool children and its effect on weight status. PEDIATRICS, 118(4), e1010–8. doi:10.1542/peds.2005-2348
Berkey et al. is my favorite because of its design.  Longitudinal cohort study to help minimize confusing reverse causal effects.  A couple quotes from that paper:

Skim and 1% milk appeared more strongly linked (per serving) to weight gain than whole or 2% milk (Table 2), although the number of children consuming whole milk was small. 

Contrary to our hypotheses, we found that (1) children who reported higher total milk intake experienced larger weight gains; (2) children who drank more 1% and skim milk had larger weight gains than those who drank smaller amounts of 1% and skim milk; (3) dietary calcium in- take was positively correlated with weight gain; and (4) dairy fat was not. The effects of milk and dietary calcium appear to be explained by energy intake, since the associations were attenuated when adjusted for energy.

O’Conner et al. is not a great paper, despite getting published in Peds. Cross-sectional analysis of some NHANES data, so it is impossible to make any reliable causal inferences.  However, even though we can’t say why they didn’t find differences, one can certainly say that they didn’t.  NB: I’m especially skeptical of this paper since it didn’t find any association between juice or soda and high BMI, either.

There was no clinically significant association between the type of milk (percentage of fat) consumed and weight status (data not shown). 

The CDC / MMWR paper is one I just found today and included just because it seems most representative of the question you were asking (but not quite).  It makes an estimate of calorie savings and changes in milk consumption when transitioning from whole to lower fat milk.  They estimate that 5-7,000 calories per year per student may have been shaved off by drinking less fatty whole milk.  However, this paper could be misleading for a couple of reasons.
  1. Their methodology skirts the real issue: that the body autoregulates and compensates with calories elsewhere.  They provide no evidence whatsoever that those calories weren’t made up for in ice cream, and the poor association between dairy fat and weight gain in other studies leads me to be suspicious that this is the case.  For example, if I instructed a patient to skip breakfast, saving ~500 kcal per day, I could theoretically estimate that they would have a 3500 kcal / week deficit… I’m sure you see where I’m going with that, so I won’t belabor the point.
  2. Even if it went uncompensated, that’s approximately 1.5-2 lb / yr of fat (theoretical kcal density of fat is something like 35-3600 kcal / lb).  In a growing kiddo.  On one hand, 2 lbs is 2 lbs.  On the other, if there were an intervention that made your patients’ lunch less delicious every weekday for a year, and the result were at best 2 lbs of weight loss… 
Anyway, I hope this makes for some enjoyable reading.  I’ll keep looking for a more rigorous analysis, perhaps something where they try to blind the kids to what kind of milk it is.
Email 2
I agree, calorie intake is the bottom line.  The interesting part is how different foods and macronutrients affect our spontaneous intake over time.  It’s tough to judge, because so much of the research is done on the acute satiating effects of foods (1-2 days, perhaps a week), but there is good evidence that our ability to pendulously adapt, compensate, and find homeostasis is remarkably robust.  
For example, “intermittent fasting” (IF) has some interesting research behind it.  It’s “common knowledge” that skipping breakfast is bad form… but unfortunately, this idea stems mostly from epidemiological / cross-sectional research that associated skipping breakfast with high BMI.  This type of research suffers from powerful confounders and sampling bias, in that people who skip breakfast generally tend to be people with overall poor dietary habits.  When IF is instituted as an intervention in a randomized sample, there are often very good results in terms of body composition / weight loss, lipids, fasting insulin, fasting glucose, etc.  
Because long-term adherence* would seem to be really unlikely on a diet that employed extended periods of fasting (often 16-24h), there has been some investigation into how the body’s hunger hormones adapt to drastic changes in eating patterns.  Sure enough, after about 2 weeks, at least ghrelin’s normal secretory pattern adapts to match when a person is “used to” eating.  Though I’ve never tried the alternate-day fasting employed in some of the research, I’ve played around with within-day 16:8 or 20:4 hour fast:feast ratios and where I only eat from 5-9 pm or so.  In my limited, N=1 experience, I’ve found this to hold true for me: after about 10 days of sticking with it, I just don’t get very hungry until my normal eating time.
It seems that one semi-exception to the “calories are king” rule is protein.  The body has very convenient storage depots for carbohydrate and triglycerides, and the number and complexity of biochemical reactions required to digest and convert the ingested forms to storable forms of these macronutrients is sufficient to expend approximately 10% or so of the ingested caloric value.  In contrast, protein does not have such an easy storage depot (darn nitrogen), and so it has to undergo a few extra steps (deamination, rearrangement, other Krebb’s cycle stuff), and these extra steps give it a relatively large “thermic effect of food” (TEF) — around 30% of the ingested caloric value.  Bray (possibly the biggest name in mainstream obesity research) et al. published the attached paper in last month’s JAMA**, which I found enormously interesting.  Here’s a quote:

With the low protein diet, more than 90% of the extra energy was stored as fat. Because there was no change in lean body mass, the 6.6% increase in total energy expenditure reflects the energy cost of storing fat and is close to the estimate of 4% to 8% for fat storage derived by Flatt.31 With the normal and high protein diets, only about 50% of the excess energy was stored as fat with most of the rest consumed (thermogenesis). The high total energy expenditure probably reflects the higher cost of protein turnover and storage. 

*As you’ve mentioned previously, adherence is really the key issue to why nearly all diets fail, and independent of whose dietary recommendations you’re using, patients / subjects tend to be back at baseline weight by 2 years.  For example, NEJM did a great, 60k+ sample size trial a year or two ago looking at low fat vs low carb vs Mediterranean diet — low carb started out head and shoulders above the others, but by 2 years there were no differences.
**Bray, G. A., Smith, S. R., de Jonge, L., Xie, H., Rood, J., Martin, C. K., Most, M., et al. (2012). Effect of dietary protein content on weight gain, energy expenditure, and body composition during overeating: a randomized controlled trial. JAMA: The Journal of the American Medical Association, 307(1), 47–55. doi:10.1001/jama.2011.1918