That doesn't make much sense, assuming there's a group of people reviewing games together. The company is going to have to buy the games in that case. The company isn't likely to care much if it's costing $20 or $50. It's nothing compared to the six figure salaries of a game rev.... wait, what? Are you telling me the Coconut Monkey lied to me about how much game reviewers make!? Well, whatever, even at minimum wage the cost of the games is trivial by comparison.
If you take it down to individuals, though, I don't think there's that much difference as long as you have an honest situation going on. The reviewer can see the price just fine and compare it to other games.
No. Reviewers are aware of how much a game costs when reviewing it, it's not like that's secret information. Reviewers also know what it feels like to spend money. Just because we get games we review for free doesn't mean we forget what buying them is like—we still buy games all the time.
It might be a factor in their value proposition, but I actually think they overcompensate for it. I've read tons of reviews where the writer talks about it not being worth the money because it was "only 20 hours" or whatever, when for many buyers I think that's a fine length and people don't often finish games anyway if you go by achievements.
I think reviewers playing games as their job, and rushing through them to cover the next game, is a much bigger factor. Reading between the lines of a lot of reviews you can see how their priority is to get through it fast and painlessly.
I agree. I occasionally read old reviews, and "Did se even play it?" is a common feeling. Civilization 5 is a good example—universally well reviewed, average 90%, while the community were mostly unimpressed. There's an opening for an operation to produce in depth reviews after weeks or months with games—not major sites or mags tho, they obviously need to catch the Day 1 traffic surge.
While I read big site reviews, the sources I trust more are Steam and Amazon. I feel there are lesser problems there once you have 1,000 or more ratings, can see the timing and scoring distribution to spot any obvious anomalies, and can filter the reviews to look at different slices—eg most recent [tells if initial problem was fixed], 1-star reviews [maybe they're all because of DRM or monetization, rather than gameplay].