Soertier, I don't think I am claiming that my TTI power supply is "Really superior in every respect". It is not superior in its purchase price, for example. Or weight. Although I will grant you that it is superior in every respect that matters to me.
I am afraid that there is a lot of misplaced opinion (in my opinion) around the facts of capacitor reliability, and many stories hark back to the debacle about ten years ago of large bad batches of caps where the electrolyte chemistry was just wrong. This caught many good equipment manufactures such as Dell out. We have moved on since then. Yes, there are good manufacturers, and indifferent ones. Let us not forget that every cap manufacturer also has higher spec and lower spec components in their ranges, and then every equipment manufacturer who has a budget to work to, and maybe makes good or poor design choices as a result. The manufacturers making the poor design decisions end up in Dave's Dumpster.
Most (but not all) capacitor failures that we are reading about here are when the caps are under the stress of SMPS switching circuits. Caps in linear supplies generally have an easier life. And even in the highly unlikely event that they ever gave out, anyone with enough knowledge to own a power supply in the first place should know enough to fix them.
There are reputable companies such as Agilent, Fluke, etc. (you know the ones) who work very hard to build and then maintain their standing. Why would any manufacture risk that with cheap components? The returns/repair bill alone would rapidly outstrip the costs of the correct (not necessarily the best) choice of component in the first place.
Look at what TTI Thandar did when the "broken clips issue" came up. How was that for good service? Yes, they cocked up with a poorly made mechanical part, which only showed up following the effects of probably rough shipping. But they reacted, and put it right. Do you think you would get the same attention from a far-eastern no-name supplier who does not have a reputation to maintain, and has achieved his end goal when his product shipped to you? I rather doubt it.
Sorry if I am sounding a bit like a Thurlby Thandar evangelist! I'm not - and I have no connection with the company. But don't let one issue with a broken component (that was fixed), or misgivings about their choice of capacitors colour your judgement. I don't believe that they made a "not so good choice" of capacitors in these power supplies.
At the end of the day, you do generally get what you pay for, especially if you take the long view. If you don't need it, or want to pay for reputation/support/performance, then through the wonders of eBay and cheap international shipping, then you don't have to. Cheap options do have their place, and it all goes back to the original question "It depends what you want it for."
Teardown? Well, I can't be bothered with lots of pics and words, but here's what's inside my PL330TP that marks it out;
Passive cooling design. Massive heat sinks, no fans
Multiple output devices. SIX (I think) 2N3055s share the brief to produce up to 3A. Some cheaper designs might risk one or two devices only
Modular assemblies on good quality glass fibre boards, all components marked
Neat wiring looms between boards. No cables flapping about
All through hole construction (but my PL330TP is no longer a current product)
Documentation! Ask TTI nicely and they may even make available calibration manuals with with full circuit diagrams and detailed descriptions
One FAT transformer for EACH power supply. Some manufacturers would get by with just one across multiple outputs
Outputs are reverse bias protected (and the 5V output even has a built in over-voltage crowbar safety device)
Good quality V and I pots, with sensible ranges and sensitivities.
Did I mention this thing is bloody heavy?
Individual output on/off switches, as well as mains on/off. When an output is off, you can set the max/constant current limit
Very good transient load response. Difficult to catch the output out in you suddenly load or unload it.
Good resolution and accuracy meters down to 10mV and 1mA. That is quite rare, and well worth having (in my view). Many supplies (including others in TTI's own range) stop at 0.1V and 10mA resolution
Fast responding meters. My displays are screaming-clear half-inch red LEDs. I prefer these to LCDs but that's my own preference.
Analogue preset pots allow internal recalibration if required (which it does not, in my case - mine is spot on)
Good quality, dual (now, that's really useful) 4mm sockets and bare-wire screw terminal output connections
Output noise is below my ability to measure accurately, at probably less than 1mV rms
Output AC mains ripple is below my ability to measure (lost in the noise)
All-metal case construction. Later modules starting to use quite nice-looking plastic moulded components
Mine has a captive mains lead, not an IEC connector, which maybe newer products have. But then it's hardly portable, so once installed, the captive mains lead is no real issue.
Maybe starting to look a bit dated, but then it is a power supply and not a fashion statement. New ones are maybe a bit cuter.
A solid, very well behaved, box.
A good number crop up on eBay, but you might have to pick your moment if you want a bargain. These are not cheap, and there's a reason for that. Still, cheaper than an Agilent/HP equiv, and there's a reason for that, too!