I think stroke number 1 in 頃 and stroke number 4 in 階 has a wrong direction. Skritter says it should be written from right, but according to http://kanji.sljfaq.org/kanjivg/ the stroke should be written from left in both characters.
The page http://tangorin.com/kanji/%E9%9A%8E also confirms it for 階 but has no animation for 頃. All other kanji with the 匕 component to the left should probably also be written with the stroke starting from left.

I asked Maksym and he was looking up the official stroke order standard for this, which he remembers saying that there are two acceptable ways to do it. This is similar to what it is in Chinese, then. We don't have a way to accept either order, so we just have to pick one.

Maksym says that point 4 - 5 of this document talks about the distinction:
http://www.shosha.kokugo.juen.ac.jp/oshiki/ronbun/cp96/cp96.htm

In particular, this chart shows that the way we have it now is currently a little more popular in 北, which is an example of this radical, but that really both are acceptable:
http://www.shosha.kokugo.juen.ac.jp/oshiki/ronbun/cp96/t19.gif

So usually I just go by what the fonts in question look like. For these two, the fonts we have agree that it looks more like a horizontal guy, so I've changed them according to your suggestion.

According to the latest 常用漢字表 http://www.bunka.go.jp/kokugo_nihongo/pdf/jouyoukanjihyou_h22.pdf , issued at 平成二十二年十一月三十日 (just 3 days ago), that stroke should go from right to left. But the accompanying notes (page 4-10, specifically page 8) also state that both directions are acceptable.

Looking at the table more carefully, 階indeed goes from left to right. Also are 比批昆鹿皆. Only 頃 and 混 are the exceptions. But even for the 比昆鹿皆’s, when they appears on the third and fourth columns, they are also shown as going from right to left. So, the direction is not so important, as even the government document that tries to clarify it shows both forms.

Ok. Thanks for changing it.