Mixing sub sizes seems to be a popular topic around here again lately. Barring the use of a relatively complex crossover network that separated each speaker's frequency response (like the 13.5 runs 20-50hz, the 10 runs 50-80hz, with the proper slopes), then no, its generally a bad idea to mix different sub sizes. Different sized subs, even of the same make/model, will have different frequency responses. Trying to couple two subs who do not react exactly the same means over all frequency response will be erratic and unpredictable. Some freqs will be artificially boosted by sound wave reinforcement, other freqs will be artificially diminished due to cancellation. This is bad for sound quality as we generally want a smooth and predictable response to make tuning the system easier.
misfit is correct in that there is no particular sound associated with a particular speaker diameter. 10's dont sound a certain way, while 15's sound different. As I said above, different sized subs, even of the same make/model, will have different freq responses. But that does not mean 15's will always have a lower Fs than 10's. A 10" sub can be (and is) built with a lower Fs than many 15" subs, as that mainly comes down to suspension design. Consequently, its not unheard of in the home audio world to have a 15" speaker that plays full-range. So clearly cone diameter alone is not the sole factor in frequency response. Larger cones do mean more air displacement, which does mean more ability for greater output in the lowest frequencies, but mating enough smaller drivers to reach the same cone area (and excursion) will give the same potential for air displacement and low freq output (given similar Fs specs).
If anything, the 13.5 has the potential for better sound quality than the 10 because due to having (a lot) more cone area, the cone does not have to move as far to displace the same amount of air and reach the same output level as the 10 would need to. Less cone excursion to reach the same output level means the coil travels less distance outside the gap, which means less BL distortion from loss of motor control. W7's are BL optimized so this is not as big of a deal as it would be for more conventional motor topologies, but its still worth mentioning.