The idea is that if you don't have a mountain nearby, you have very little to work with (the roads UA by itself is usually worse than papermaker since there's one less road than # of cities). ense7en, what were your exact stats on a fractal/continents map for capital by mountain, second city by mountain, and not really any mountains for the starts you rolled? I'm curious, and "mostly" is kind of vague. I think a no start bias roll is something like 20% on a fractal map for viable mountain placement next to your capital, less for continents, more for earth and pangaea. If the hills start bias is really pushing the start to something like 50% then I'm moving both Incan and Austria up one each, since that +30% chance for mountain start would be a pretty big benefit in and of itself.
My general theory in ranking the civs: If Inca gets graded an A for 50% of games, a B for 15% of games and a F for 35% of games; Siam gets graded a B for 90% of games and a F for 10% of games.... I would not hesitate to rank Siam as a clearly better civ than Inca. To take an example, goodolarchie mentions the great game he had with Inca, on his SECOND roll. You would almost never say that about any of the top or upper tier civs... because almost all starts are good starts for those civs. The idea isn't "look how great Inca is when they have mountains", no one doubts that they're a higher tier relatively if you reroll all civs until you get a mountain start for each of them... you just have to modify that enthusiasm by the % of times you actually do get mountains.
In any case, enough people have spoken up. If no one backs me up and one more deity player thinks Inca should be moved up, I'll move them. Also, I think at this point the only debate is what% of Incan starts on balanced maps actually has mountains, which can be resolved statistically. No one seriously thinks Incan are anywhere near top tier without mountains, they'd be solidly lower-mid tier, very similar to Persia (better movement speed for military and has culture/hammer/gold split instead of pure gold bonus; which is fairly comparable). If the hills start bias really lets you have a mountain start 50%+ of the time, then I'd move both Austria and Inca up.