the dealy of 4-2 compressor in multiplier

related:www.patentstorm.us/patents/6308195.html

the delay from the input of 4-2 compressor to the carry out port is equivalent with the delay of a 1-bit full adder.The carry out of lower bit 4-2 compressor is connected to the carry in of upper compressor,which is just like the Carry Ripple Adder.
Does it mean that the delay of a 32bit 4-2 compressor is as much as the delay of a 32bit Carry ripple adder?
Usually there are 2 stage 4-2 compressors in a 16-bit multiplier,and 3 stage 4-2 compressors in a 32-bit multiplier.The delay of these stages seems to be too large ,if the answer of the above question is yes.
Is there another explanation?

the answer is no,
because the current output carry signal is only depend on the current input data, it has no relation with the carry signal of the privious level, (maybe you can write out the logic express to check)
that is, the delay of 4-2 compressor is almost independ with the width of the 4-2 compressor. for example, the delay of 32-bit 4-2 compressor and the delay of 16-bit 4-2 compressor are almost the same.
while for carry ripple adder, the delay is depend on the width
that is why we use 4-2 compressor when design multimplier

Added after 6 minutes:

by the way, the logic expression of 4-2 compressor is not only one kind, there are many implemention method, the key is the output carry is only depend on the inpout data, and has nothing to do with the carry of the privious level,

tarkyss said it correct. you can refer to this link
https://www.geoffknagge.com/fyp/carrysave.shtml#42