I agree with KV and in addition to sample to sample variation it will also be affected by how well the individual lens is calibrated to the AF system of the body, how it is typically used by the owner (i.e. does he/she typically bias the exposure triangle to shallow or deep DoF), and individual tastes.
It is probably more useful to choose instead based upon avoiding weak points that may be important to your work (flare, form of distortion, focusing speed, etc.). I take a look at several reviews before buying new glass and I use them largely like I do reviews on the Amazon site; I look for faults in particular and decide whether it is likely they would be important to my work and also whether those faults are user induced or exacerbated. And ultimately there is no perfect choice because every lens has benefits and drawbacks. I went through this recently because I really wanted to decide on the Canon 200-400 1.4X F4/F5.6 lens which is extremely versatile. There are many cases where its ability to handle a range of sports shooting situations would be really helpful but I ultimately concluded that given the light conditions I often face, the less versatile EF 400 F2.8 is the better choice for me. There will be cases where the EF 400 can get a shot the 200-400 cannot but the reverse is also true. Ultimately it comes down to maximizing the performance of the glass you have, modern cameras are "intelligent" but the owner should be more so in order to get the most out of the system.
And it is important to avoid the tendency to "pixel peep" looking for issues and aberrations which are of no importance in a normal size photo either printed or web displayed. I have lab grade electronics test gear that can measure waveform distortion down to a vanishingly tiny amount so it can be used to distinguish between two excellent pieces of home audio equipment however in use no listener would be able to discern that difference in a blind test no matter how carefully they listen. A lot of the microscopic analysis of photos falls into the same category and time would be far better spent on composition which makes a huge difference rather than deciding that there is a minute flaw at the discrete pixel level in a corner of a photograph.
Rodger
EOS 1DX M3, 1DX M2, 1DX, 5DS R, M6 Mark II, 1D M2, EOS 650 (film), many lenses, XF400 video
