the seasonal variation argument doesn't seem to be as strong as many say.  According to this site demand varies from a low of 8.8 million gallons per day to a high of 9.6 MGPD.   And if you look at the chart for production of gasoline it strongly correlates to demand.   Seems like they have no problem meeting demand.  So if production can easily meet demand, why would that change prices?  It wouldn't in any other industry.   The "people use more, so we'll make more and thus charge more" model doesn't usually apply.  

Just one more thing that doesn't add up.

If you look at the way that US commercial crude oil inventories have been declining for the past 4 months or so, it's clear that US refiners have been "betting" that the current surge in oil prices is a temporary spike that will reverse in the near term. If you do the comparison on a same-week basis, spot crude prices are 50% higher than the end of October last year, whilst product prices are 25-30% higher; however, if you bear in mind that the oil that is currently being refined is probably $75-80 oil that was purchased 2-6 weeks ago, then the price rises are actually fairly consistent.

FWIW, I doubt that crude prices will moderate significantly as inventories are likely to post further declines this week and next on the back of weather-related outages in Mexico - and that US gasoline prices are likely to rise some 30-40 cents in the none-too-distant future.

Given that US refiners had $25-30 crack spreads earlier in the year, they can easily absorb a period of $5 crack spreads if they believe that this will be a temporary - ie less than 90 day - phenomenon.