I assume (I have no doubt you are very knowledgeable in this field) you made a typo or something, as the sentence "10w to listen at 89 dB 1W/1m and 100 watts........." seems confusing. Based on the sensitivity spec of 86 dB/2.83V/1m as specified by Wharfedale, it would be 2W to listen at 89dB at 1m if the nominal impedance is 8 ohms, or 4W to listen to 89 dB at 1m if the nominal impedance is 4 ohms. So to get 92 dB at 1m, only 8W would be needed, not 100W, and that's for just 1 speaker. The specs also say "Peak SPL: 95dB", presumably at 1m, that would only require 16W even if the nominal impedance is 4 ohms and again, that's if only 1 speaker is playing. I think a 100WPC amp will be all he needs, but even a real 50WPC amp will give him plenty of headroom as it is obvious those speakers are not designed to play loud or take too much power.
To increase the acoustic output by 3dB you need 2x the amplifier power in watts. However you also need about 13dB of headroom above the average power level to insure no distortion on transients. The sine wave used for measuring sensitivity of loudspeakers is in essence a continuous or average power level, without consideration of transient signals.
My suggested power levels are just the levels I found desirable with Solid State amps and those measured average SPLs to accommodate clean transients. It wasn't intended to reflect the strict average power level required to power the average sensitivity.
Your clarification is welcome; I didn't intend to mislead by implying the values were strict measured values. Just practical values in real-world listening.
Modern amps tend to have less robust power supplies than older ones, where big transformers were the norm and there was no competitive advantage in using smaller VA rating units. When you're shipping from China instead of across one or a dozen US states, weight looms large in the retail price. Today there is a huge incentive, because retail prices are such an important part of marketing in the internet age, and the consumer has a large number of tools to read the market price trends. So even $10 can play a large role in overall sales.
The result is that transient power ability is compromised; not necessarily with just a single tone burst as the supply can usually deal with those if not superbly, at least adequately (as is done when creating amp specifications) but repeated transients over a short period of time. The remedy is to seek higher average power output.
A related issue is as the loudspeaker is driven, some power is lost as heat. There is limited ability to dissipate heat in a driver so that heat tends to accumulate. As the power is increased, so does the heat generated in the voice coil. This has the effect of raising the loudspeaker's impedance, lowering SPL, and encouraging turning up of the amp, which adds more heat, which raises impedance, which lowers SPL ... well, you get the idea.
Totally out of the scope of this particular discussion, but perhaps worth mentioning is that vacuum tube amps seem capable of repeated short-term transient power considerably beyond their average power ratings while still meeting the average power rated distortion, while with a transistor amp if you can manage 2x, you're doing quite well, and 1.4x is quite common. It's a rarely explored area of the power envelope, but what little has been done supports it.