Yeah it really goes against everything i have ever read or been told. US-05 is known for its fairly poor flocculation, which results in the yeast dropping slower. You want yeast kept in suspension at the end of fermentation to clean up.
I can only gather that they rely on the yeasts poor flocking and reluctance to drop out of suspension lets you get away with lowering the temp.
But if you were brewing a faux lager type ale at 15 - 16 degrees you may have had a slow start to the ferment at these temps and you would definitely want the temp to rise a few degrees at the end to help keep the yeast active.
I always let all my ales rise a couple of degrees at or near the end of fermentation then leave them a few days to a week before dropping the temps for ccing, often the beer clears on its own at ferment temps.
Here is some further reading from the Blue Book on yeast from the IBD.
"The reduction of vicinal diketones in the later stages of fermentation and during maturation
requires the presence of adequate yeast in suspension in the fermented wort. Thus, where
the yeast is particularly flocculent (this phenomenon will be discussed later), premature
separation will be reflected by low rates of diacetyl reduction and potentially elevated
levels in finished beer. Diacetyl removal is also affected by the physiological condition of
53
the yeast. When the pitching yeast is in poor condition, such that the primary fermentation
performance is suboptimal, the yeast present during the latter stages will be stressed and
the period of diacetyl reduction will be prolonged.
A number of strategies can be adopted to ensure that beer diacetyl specifications arc
achieved. Diacetyl removal can be attained post-fermentation in the conditioning stages of
brewing (traditional lagering). This is a slow process, expensive in terms of time and
conditioning capacity. Alternatively, it is desirable to ensure that minimum diacetyl
concentrations are achieved before the beer is removed from the fermenter. It is necessary
to select fermentation conditions (i.e. pitching rate, wort DO and attemperation regimes)
which provide an optimum profile. In practice, the aim is to promote the maximum aacetolactate
levels as early as possible, such that the resultant diacetyl may be rapidly
reduced due to the presence of a high suspended yeast count. This reductive phase may be
stimulated by increasing the fermentation temperature approximately two-thirds through
the fermentation cycle. "
