1. No, the correct answer is that the question cannot be answered. The question is effectively this: If X = 5 + ? + ? and Y = 4 + ? + ? which is greater? This question is impossible to answer without knowing what any of the question marks represent. Level, including compression, is just one of the parts of the equation our brains use to calculate loudness. In my example we know the level, represented by the 5 and the 4 in the equation but not the other variables.
Here's another example but with those variables filled in: X is a signal comprised of low frequencies which has been compressed by 6dB (has had it's peak reduced and 6dB of make-up gain applied) with a peak value of -0.5dBFS. Y is a signal comprised of mid-high frequencies which has been compressed by 3dB with the same peak value of -0.5dBFS. Assuming both X and Y are roughly the same duration, which is louder? ... We can now answer the question and that answer would be Y, by a huge margin! The peak levels are the same but the RMS level of X is greater and therefore X should sound louder but it will actually sound very significantly quieter.
I created this last example with an additional point in mind: Let's say we have a piece of music which is comprised mostly of our X and Y signals, what would that piece of music look like in your DAW? ... It would look like a sausage because both X and Y have peak values which have been compressed/limited to -0.5dBFS, but because Y sounds many times louder than X, our piece of music has a large dynamic range!!
2. You seem to be arguing against yourself, without even realising it! You keep going on about peak levels, that music looks like a sausage in your DAW due to compressed/limited peaks, your avatar is even a picture of it, but if peak levels "have a negligible effect on loudness" then how does looking like a sausage have any more than a "negligible affect" on loudness or dynamic range??
G