Vegelen, since I'm avoiding doing actual work for a bit I'll attempt a start of an answer to your question, and tbh if you want anything more detailed you'll need to register for a master course if you can find any providers who actually know what they're talking about (most don't in my experience once you get into the details).
There is no one size fits all optimum panel and inverter combination, there are many different optimum combinations depending on the characteristics of the site, requirements of the customer (eg do they want a decent wireless monitoring system, do they want black framed panels for the looks), and the costs and availability of specific equipment at that point in time.
In general if it's a roof with limited space we'd usually prefer the Sanyo panels both for their increased generation capacity for that given roof area, and their higher operating voltage to increase the inverter efficiency, as well as ~5% performance improvements from their better high temperature and low light performance. This used to be a no brainer for anyone with the money, as the improved FIT payments outweighed the increased costs in percentage terms most of the time. It's now no longer so clear cut as the costs of the cheaper panels have plummeted so much faster than the Sanyo panels that they probably will give slightly better returns in percentage terms.
If there's shading we'd generally use the 4000TL (now 3600TL) due to its dual MPPT and optitrac global peak shading function. Other options could be solar Edge, or micro inverters, which both have pros and cons.
TBH we generally specify the 4000TL anyway for most of the 4kW jobs because most want the sunny beam option, although on unshaded systems I'd think that either the aurora or fronius inverters would be better in single string mode when matched with 16 x 30V panels due to their higher voltage range allowing the single string which is going to be closer to their peak efficiency range. If the option is there, we'd advise on the use of eg 20 x 190Wp 36V panels in place of 16 x 240Wp 30V panels though because the string voltage is closer to the optimal, but this isn't always practical / cost effective.
For big roofs with no shading though, it really is very hard to argue against installing basic cheap chinese panels to get a 4kWp array up as cheaply as possible, as the overall cost savings are almost certain to far outweigh the potential increased financial returns from the better panels (IMO usually in the 1-3% range if anything, unless you've made a really bad choice for the cheap panels). Having installed a lot of panels from a lot of manufacturers, I'm pretty certain that price doesn't always equal quality, and that many of the cheap chinese panels are as well built or better than many higher priced European panels. They're also almost certainly built in brand new factories using the exact same machinery as most of the more well known Chinese and other brands, and the kit they use would probably put Sharp's Wrexham plant to shame if Sharps publicity shots are anything to go by.
There are legitimate arguments about if they'll be around to honour warranties etc, but tbh other than the major brand names that have massive interests outside of solar, I doubt there can be any guarantees about any of the big solar companies (which is why we like using Hyundai, Sanyo / Panasonic etc.). I remember a time when Solarex were about the biggest solar panel manufacturer in the USA, then got taken over by BP, became BP Solarex, then BP dropped the solarex name, failed to invest in it properly, massively dropped market share and eventually closed its solar business last year, before apparently restarting it again this year based on buying dirt cheap chinese manufactured panels and rebranding them as BP...... personally I reckon it's more honest to simply buy and sell the dirt cheap chinese panels directly instead of selling them badged as BP at a premium... but I digress.
I'm also of the opinion that essentially, as long as the build quality is there, then one mono or poly crystaline panel is basically much the same as another, and the degradation rate isn't going to be a lot different between them unless they use seriously mismatched cells, or dodgy bypass diodes.
IMO the place not to scrimp really is the inverter unless roof space is so limited that the efficiency of the panels is actually a factor, as cheap inverters will almost certainly have cheap capacitors and other components that will fail much earlier than those using higher quality components. In Mozzers case I believe he's got 2 x 4000TL, which definitely doesn't count as scrimping.
The key to maximising performance to value ratio IMO really is trying to best match the voltage and capacity of the array with the inverters peak operating efficiency point, which is why comparing different panels on different inverters with different string set ups is such a poor way of comparing actual panel performance.
right, I'm bored now so I'll leave you to ponder on it yourself.
