I have two Leica confocals with WLLs (NKT Photonics), both acquired with the systems through Leica. In my experience, there are only a few 'cons' --
Initial costs: A WLL can add $100K - $200K to the cost of a full system, when compared to a system with fixed-wavelength diode lasers. However, this is not so different from adding a laser for multiphoton fluorescence. Both lasers are bundled into the confocal system service contract, so I am unable to give you numbers on a service agreement for the laser in isolation.
Repair turnaround times: In the rare instance (twice in ~9 years) that the WLL needs factory repairs, the turnaround time can be long (weeks - months). In my case, this was not overly painful, as a loaner WLL was available.
Somewhat lower output power at selected wavelengths: This is less of an issue with newer versions. While diodes can provide more mW at their specific wavelengths, this is rarely an issue for biological specimens which can withstand only a fraction of the maximum available excitation laser power in any case. If higher powers are required, e.g. for photobleaching, applying e.g. 512, 514, 516 nm simultaneously delivers more than enough energy.
Tuning range: Again, less of an issue with newer versions. On my older system I have 405nm and 445nm diodes in addition to the WLL (470-670nm); on the newer one only the 405 nm diode is required (WLL 440-830nm).
A sort of 'global con' is that the rest of the confocal system has to be able to take advantage of the various capabilities and characteristics (tunability, pulse synchronization, etc.) of the WLL to make the purchase really worth the price, especially if you will be adding only the laser. What kind of confocal would you be retrofitting?
An argument can be made that having the ability to output 517nm instead of 514nm is a luxury (although I did have to do exactly that for a recent imaging experiment). However, if your users bring you uncharacterized fluors, or unusual fluor combinations, it's very hard to go back to fixed laser lines after using a WLL.
If there's a specific drawback of concern, or any specific questions, I'll be happy to elaborate as best I can...
Mahmud
Mahmud Siddiqi, PhD
Imaging Facility Manager
Carnegie Institution / Embryology