In case you haven't noticed before, I have -- since gross dissatisfaction and disillusionment with the mainstream "reviews" of them -- done my own, independent research (both theoretical and empirical) on portable heaters, quickly gaining a whole new appreciation (and, to a fair extent, love even) for them.

So, because it's an excellent example of my point-to-come, I'll start with an overview of what's in a portable heater (the regulars here doubtless already know this, but it's for the convenience of newcomers):

(For argument's sake: I'm considering only compliant products to modern standards, as best as I understand them, at least in the 'Western' countries i.e. Australia, Europe, USA/Canada etc.; non-compliant products, as may be imported direct from China -- but are also found, to an increasingly scary extent, in even "reputable" retail stores shocked -- are irrelevant to the basic arguments.)

Sure, the current draw is "relatively high" (only by comparison to other mains-voltage home appliances, though); but given that pretty much all electrical parts are sized according to the expected loading, that's largely irrelevant in the overall scheme of things. (Because of tensile strength and fault current considerations, mains cords aren't allowed to be much smaller than needed for 8--13A under optimal conditions; that's all.)

The heating elements themselves are just high-power resistors in specialized forms. Resistors in general are reliable components when properly manufactured and operated within their ratings, and these are no exception; and when they do fail, it's generally as an open-circuit, not catastrophic (unless subjected to an insane overload, causing a meltdown).

If you look at a radiant heater (that is, a true radiator with the highly visible glowing elements in front of a reflector; these are, by the way, the only type really capable of igniting objects external to the heater itself, in normal operation), it's pretty much just the heating elements, a few power switches (+ maybe a thermostat on 'fancier' versions, not that I find binary thermostats very useful), a tilt switch in most (at least modern) units to cut the power off if knocked over, and perhaps a bimetallic switch to protect against overheating (of the heater itself, anyway).

If you look inside a convector heater (of the 'plain' type, with bare elements in free air), there's very little electrical difference. The elements themselves, operating at somewhat lower temperatures than in a radiator, should also last much longer (their lifespan increases exponentially as the temperature is reduced -- as is a prevalent trend in thermal engineering). A slight hazard still exists if something flammable makes its way inside and onto the elements, but this can be mostly guarded against with suitably fine grilles; otherwise, convection heaters have a huge safety advantage over radiators.
These always have the bimetallic switch (usually mounted shortly below the heating elements; that position is rather suboptimal in practice, but still far better than no protection), so that blocking the airflow won't (in theory) start a fire. Some have a thermal fuse on top of that, as a last-resort measure in case the bimetallic switch sticks on; unfortunately, because thermal fuses will (not entirely unlike load fuses) degrade rapidly and cut-off prematurely if run too close to their limit, many manufacturers seem to have ditched them (it's very likely that most buyers' complaints guessing at element failure, are in truth caused by the thermal fuse blowing).

If you look inside a fan heater, there's still not that much overall difference apart from the fan itself (and that the thermal fuse is mandatory, in order to permanently disable the heater when the fan stops). (The plastic casings common on these can indeed melt if they get too hot; but this should be preventable with a sufficiently robust plastic, and correct temperature setting of the thermal fuse.)
By the way, the lifespan of the fan bearings (the main limiting factor in competent units) can be extended almost indefinitely with a bit of appropriate bearing oil in each, every now and then.

Indeed, even after all the penny-pinching, I still find convection heaters from half-decent companies (i.e. those that actually engineer their own products, rather than just slapping their brand onto generic clones) to be among the most reliable classes of household appliances! (Comparing appliances of similar age, at least.)

That established: If we really couldn't make a reliable space heater, then how could we have any hope whatsoever for anything more sophisticated? crazy

Now, accepted "wisdom" in the industry (or much of it anyway) seems to be that crimped connections are more reliable than solder. For high quality crimps, sure; but most that I (and presumably many of you) have seen in recent equipment aren't nearly that good.
The bane of low-end crimped connections is that although they (usually) run cool enough to start with (thereby passing initial tests), corrosion develops in the wire-to-terminal interface (at a rate dependent on temperature and atmospheric conditions) which causes a gradual increase in connection resistance, causing more heat, and ultimately a vicious cycle. (Good crimp connections mitigate this, by providing as much protected contact area as possible.)

From most reliable to least reliable, I would rank connection types (apart from screw terminals) in about the following order:
  • Spot weld
  • High-quality crimp
  • Leaded solder (R.I.P. frown )
  • 'Also-ran' crimp
  • Lead-free "solder" (I'm not going there in my own projects, unless I somehow have no alternative)

(Even then, I'm not sure every lead-free solder would be necessarily worse than 'also-ran' crimps under substantial load.)

I've also done full-load testing of my Arista SPB 1 (6-way power-board, one of the last 'good' ones made), which has internal connections made with leaded solder, and they stayed cool to the touch (only at the breaker/switch did they get noticeably warm, but that's fine allowing for the thermal trip element). (Intuitively, any competent solder connection shouldn't internally corrode in a hurry; as long as it stays below the melting point of 183°C for 63/37 Sn-Pb alloy, it should last very well.)

Given that I own neither a spot-welder, nor high-quality crimp tooling; that leaves leaded solder as the best available option for permanent terminations in my own devices, so that's what I use.
(By the way, I know very well not to solder wires that will go into pressure connections -- be they crimped or screwed; and I try to avoid it even in spring terminals.)

This came to my mind after having proven before that flexible cords can handle much more current under optimal conditions than many (especially British) regulations give them credit for. Additionally, it seems to be not only properly fitted rewireable connectors that run cooler, but also older cordsets (from the days when they were made in Australia, Europe, the UK, etc. cool). (After all, if it was really the wire size itself that was insufficient -- supposing for argument's sake that they got it wrong in the first place -- how could it have possibly taken 50+ years for the VDE to "notice"? The irony here is that their ineffective "shotgun" remedy almost certainly costs more in total, than making good connections in the first place ever did. grin)

Like many regulars here (past and present), I definitely agree that it's very often (moreso now than ever before) safer in practice to just keep using products that have passed the test-of-time; rather than gambling on new ones.

In the end, though, I suppose it's yet another symptom of consumerism (the epic disaster that's been). (If the public didn't become so tremendously wasteful, there would be no need to penny-pinch on such basic items.)
And consumerism does seem to be cruelest to mature technologies (as once companies run out of worthwhile innovations, they resort to gimmicks; and when they run out of marketable gimmicks, price is about the only other thing they can "differentiate" themselves on).

To change the game, then, I'm working on a plan to make proper reviews (of appliances in general, but especially portable heaters); while I have yet to get all the required test equipment, you can see some preliminary work (including two contrasting fan heaters) in this thread (which also includes an occasional non-electrical item, where I see fit) on Hardware Insights (where I'm a forum moderator, and occasional publication editor).
(I'm quite convinced, by the way, that the "flat" fan heaters exist largely to pander to the improper reviewers...)