I was intrigued by it, and had high hopes. I thought the previews on the Mongoose site looked interesting, but the discussions on their forums suggested that there may be a few problems with the way that some of the rules have been implemented (there is one especially silly rule about handling contests between people when skill scores for one party are over 100% - a rule which essentially reduces your chance of success for having a higher score!).
I went into my FLGS today to have a detailed look, expecting to buy it - but I didn't in the end. Looking through it reminded me just how good the 2e RQ was when I used to play it, and didn't actually thrill me with new possibilities - it just reminded me how good the old one was.
I'm sorry that the basic book doesn't come with more Gloranthan specific stuff in it - I understand that will be a separate book.
It seems to me that the best way of handling skill contests when one or more people have scores of > 100% would be to borrow from the Heroquest rule idea of mastery, and the person who gets the best result wins.
e.g. trollkin with 30% gamble against lucky fateman, with 140% gamble.
Trollkin attempts his 30% and will either crit, succeed, fail or fumble.
Lucky Fateman takes off 100 as a level of mastery, and attempts a 40% check, but his level of mastery automatically bumps his result up one, so he will get either a crit, a succeed or a fail, but can't fumble. (if he rolls 40 or less, his success is bumped up to a crit, etc).
If Lucky Fateman finds he is rolling dice against the god of luck and death, who has 370% gamble, then to start with we knock masteries off both sides until someone is out of masteries; after the first go Lucky is down to 40% and GoLaD is still 270%. This means that GoLaD has two masteries in hand, and he makes a check to get 70% or better, but his result is automatically bumped by two levels (so if he rolls a fail, it is bumped up to a crit, if he rolls a fumble, it is bumped up to a success). Lucky might roll 01 and GoLaD might roll 00, but Fatemans chances don't look good!
(Lucky vs GoLaD is just an exercise - you wouldn't really expect such a mismatched opposed check to actually appear in the game. using the Mongoose RQ rules you would halve both scores to 70 and 185, then halve them again to 35 and 93 - which gives a relatively huge chance of success to Lucky. But then again the trollkin would still be on 7%

)
Cheers