Eh, not a very good review. This whole D20 derivative aspect is wrong (not too hard to see where that comes from though) and not especially helpful. WFRP obviously derived a lot from the original D&D, but its inheritance from 3E is paltry, no more than the level of derivation than many have seen in 3E from the 1st ed WFRP.
In the setting description, it also seems to have completely sailed over the author's head that "the Empire" is a reflection of the Holy Roman Empire, not "Germany."
The example demonstrating how the mechanics of the system are "identical" to D20's is highly questionable:
Example:
A WFRP character who wanted to try and hear a noise would roll 2d10, and try to roll under their relevant stat percentile value. A D20 character who wanted to attempt this task would roll 1d20 and add a bonus vs. a DC of 20. Assume the stat in question was 35% and no external modifiers apply. In WFRP a 2d10 roll of 34-01 would succeed. In D20, a 1d20 roll with a +7 modifier (35% divided by 5) vs. a DC of 20 would succeed on a roll of 20-13. The results are mechanically identical.
Yeah,
IF the stat in question is 35%, then that's true, but do I have to point out the obvious, that in a percentile system, that only happens a fifth of the time? These are not mathematically identical mechanics.
The example is also mistaken, in that WFRP skill tests don't always go off the raw ability stat - you do often add in a skill stat, in the above case Perception.
If the author had understood the system better, the author would also have realized that one of the key differences between D20 and WFRP mechanics is the higher granularity and more limited dynamic range of WFRP stats. In d20, there's no limit to how high your skill check can get - in WFRP, a 35% Int stat (which is VERY good) isn't ever going to get higher than 70 (fully advanced Wizard Lord). You could, with a great deal of effort, take the Perception skill 3 times to get that up to 100, and with the Excellent Hearing talent you could get to 120, which still gives you a 10% chance to fail on a Very Hard test. Almost no Warhammer character will ever even come close to that level.
Additionally, what happens in practice with the percentile stat is that it usually creates at least a tiny little gap between the character and perfection, or between the character and absolute incompetence. Exactly because the bonuses, threshholds, and penalties are all on 5% intervals (for the most part, only the initial base abilities use the 1% steps), it very commonly ends up that the character will try something with a 97% chance of doing it (or a 7% chance). In D20, there's much more of an all-or-nothing effect, just because it doesn't have that little bit of odd slop in the WFRP percentiles. Of course, you wouldn't tend to appreciate this without playing the game for a while, but then if you haven't played the game, why are you writing a review of it?
WFRP does not have stand-alone saving throws. Instead, effects that in D20 would require a save are explicitly bound to ability score checks. Bonuses or penalties may be applied by the effect or the GM depending on circumstances. This is similar to D20's system for ability-score linked saving throws.
This also displays the author's bias. WFRP does not have saving throws at all. Period. It is not a concept in the game, and yet he imposes the concept on the system merely to show us that if one
could identify a saving throw equivalent it would be similar to the D20 system. So... the idea is that rolls based on abilities, potentially with GM imposed bonuses and penalties, is a similarity to D20. I guess pretty much all gaming systems are basically just D20 with cosmetic alterations, when you get right down to it.
The overall impression I get from this review is that it's mostly Ryan Dancey's soapbox to convince us that D20 is the font of all gaming. It's somewhat interesting just to see what a D20 designer thinks of game design, but on the subject of WFRP, it's neither especially insightful nor accurate.
**OOO
. Here are the direct links to the 
