Hi Ened,
No essential difference. The one-step is but a template that is reused over a multiple-step binomial. Either one or two-step is unrealistic, both are (in John Hull) for learning purposes.
But one thing I would offer is that the one-step holds the key, i think, to understanding the elusive RISK-NEUTRAL VALUATION (try finding somewhere a good explain of this!). Here is snapshot of the slide that mimics John Hull’s example of a one-step, where the stock S = $10 and can go up (S)(U) to $12 or down (S)(D) to $8. This is similar to Hull’s 11.2 where S = $20, goes to $22 or $18:
The instructive point of this is (i) to show the approach is based on a riskless portfolio [i.e., long delta shares plus short the call option is indifferent to the outcome; so portfolio is riskless]. To understand this is to understand something counterintuitive: the probability (p) of an up step is irrelevant to the option price (but note: u and d are relevant, as volatility increases they will increase/decrease). The is the essence of the risk-neutral idea: (p) is irrelevant, so that the expected return on the stock does not matter to the value of the option. Why? A higher expected return must be met (i.e., offset by) with a higher discount rate.
David
Screencasts on binomial pricing 
