you did say Nvidia couldn't beat that score. They may very well can if they run the benchmark at the same settings.
I think Nvidia can beat it if they release a new product. The 780Ti they're coming out with is probably to take some attention away from the 290X release. That was my point about 780ti needing to be the full 110. This is really a case of newbie overclockers thinking Tess modified results are invalid. If the bot didn't allow it, then it wouldn't be on the charts.
For now the 290x has the #1 spot on the bot. If Titan can beat it, then it will. We'll see. If this 780ti has all the cores of the GK110 /180 on the B1 stepping, then there's a chance it could take back the lead. Currently according to the hwbot website, the 290 is in first place, and it's hard for Nvidia fans to accept that. Hardcore Fanboys pull the 'not valid because of tess' excuse. If someone can beat that first place 290x score with Nvidia cards, they will. They'll do it tess modified or not.
If the majority of people won't accept the score as valid with the Tesselation modified, then the 290X benchers will be forced to try to beat the Titan Quad Sli scores without modfied tess. Which may or may not be done... As it stands now, professional overclockers don't care about that setting, which is why it's being enabled and the 290x is still ranked #1 on the site in the first place. Kingpin has GTX Titan scores taking first place on Futuremarks Hall of fame (3dmark website) that are a lot LOWER than the scores he puts on HWbot. Think about that for a second.
If enough people make a fuss, then maybe we'll see some Titan tess modified, or 290x without tess modified. 290x is #1, $450 less, and hard to swallow. How important is tesselation to that benchmark? That's the real argument going on. Is it important enough to make them change the rules and not allow modified tess to give the 290x #1 spot. Is it that hard for NV fanboys to deal with a Radeon card in first place? Whats the reasoning behind the 780Ti release?