I'm a little in the dark about what the OP means by 'Know Binary'.
Is it absolutely essential to know that 0000 1011 (base 2) is equal to 11(base 10) ?
Probably not.
It is, though, a good idea to have a feeling for the numeric representations used in PLC's (and other programming). If only to recognize common tricks. But also, it is necessary to understand typical 'binary' (as in the terms, not the numbering system) operations.
In AB terms, for example, N7:20 OR 5 will set bits 0 and 2 to a 1. In any PLC, shifting an integer is the same as a multiply or divide by powers of 2.
At the very, very least, it is useful to know (in most, NOT all, cases) how to use the Windows Calculator in 'scientific' mode to convert between the most common integer systems.
Base 2 (binary) is just a simple way of representing integers (read, actually, 'Pre-defined ranges of Whole Numbers') that is 'natural' to computers.
I had to specify 'Pre-defined... yada', because I'm positive that someone will point out that integers can be virtually any length, (being defined as the 'Natural Data-Width of a Particular CPU'...).
As far as numbering systems go, in addition to 'knowing binary', you also need to be aware of the natural size of the word on your particular platform, as well as how numbers are represented.
For 'SIGNED' 16 bit integers, represented by 2's complement, the numeric range is -32768 to 32767. The exact same word, interpreted as 'UNSIGNED' represents 0 to 65535.
Also, the representation (in base 2) might at first be counter-intuitive. Same 16 bit integer, (bit-count starting at 0), bit 15 is the 'sign' bit. If true, the number is negative... So, 16 bit, signed, 2's complement, 0000 0000 0000 0001 is equal to a decimal 1. The odd part, is 1111 1111 1111 1110 {{{EDITED As per Peter's 100% accurate post below pointing out that once again, I'm in error!, Thank you Peter Nachtway... Credit where Credit is due}}} is equal to (-1) in decimal.
Why? Because using 2's complement, there is a unique representation for every possible whole number in the range. Another, older, though sometimes seen format, called 1's complement looks at first glance to be more 'natural', representing (-1) decimal as 1000 0000 0000 0001. This has a HUGE problem though.... as an exercise, let me know what the problem is