Faulty VW Polo 1.2

Hi all.
Long time, no post.
A quick summary of the problem...
52 plate 1.2 (3 cyl) 60hp engine Polo. It's done ~51k miles. It was
last serviced (AFAIK) last September, it's 6 year service.
It's got an intermittent/transient problem where the CEL light comes on, reporting a P0420 (inefficient cat) problem. When it does this, the car smells of petrol and uses a lot of fuel.
If left on its own, it'll go away after 48 hours.
When the light isn't on, then the fuel usage seems reasonable to me - I've driven it to work, 70 mile round trip, and it's used ~20% of a tank, which gives a range of 350 miles, which is about right, certainly not massively down on what I would expect.
I've replaced the coolant temperature sensor (this has happened before, and replacing the CTS fixed it). It's had (some time ago) all 3 coil packs replaced. It had a new dash pod last year (seems to be the main CAN hub - that goes south, the entire car stops working)
The idle seems to be a bit on the slow/lumpy side (750rpm, and bearing in mind its a 3 pot!), and the car seems easier to stall than normal.
I'm not convinced that the problem *is* in the cat, as the lumpy idle is when it's at idle(!), and shoudln't be running in closed loop.
I'm not keen to take it to our local garage because (to paraphrase them), "They read the codes, replace whatever it says, and move on from there". Given the code it flashes up is a weak cat, I don't want them to replace the cat (probably 500 all in), and maybe both lambda sensors (another 300 probably) for this not to be the problem.
I've got an ELM327 clone, plus some free software. The No1 O2 sensor shows "N/A", the No2 O2 sensor shows (as far as I can tell from a 3Hz sampling rate) that it's swapping between 0v and 0.9v at about 1Hz when running at part throttle on a flat road. Foot flat to the floor gives a constant 0v, foot off and coasting gives 0.99v constant.
As I understand it, if the car is being over-fuelled, it'll overwhelm the cat, and the ECU may well then see the error as being in the cat, rather than highlighting the *actual* problem.
I realise that the reason for the over-fuelling could be a leak (either in the intake or vacuum side, including crank case) downstream of the MAF. I also realise the problem could be in the upper O2 sensor - if it thinks the mix is too lean, it ups the fuel, floods the cat, which then brings up the P0420. If the sensor is weak, but still toggling, it may not be picked up by the ECU, because it's still toggling, just slowly. I don't seem to be able to get a figure from this sensor in the ODB software (just shows N/A as though it's not supported).
I also realise it could be the same thing that Honest John refers to as "random over-fuelling" on his review of this age/model/engine Polo.
So far, it's only done it on "short" journeys (i.e. about 5 minutes into a journey). Normally it's put onto a motorway and driven at 60 (which is still 3.5k RPM IIRC), but when it went on a shorter one, it then lights up.
Does anyone have any advice or suggestions for..
1) Order of checking/replacement of components? I've done the coolant sensor (wasn't it). It *should* have had new plugs at its last service, but I take it that a plug check should also be done. Is there any mileage in removing the O2 sensor and checking its physical condition? If it's sooted up (due to the problem), cleaning it up?
2) Suggestions of other places to check like "9 times out of 10 it's <this, that or the other>" I realise that lambda sensors have a life of about 50k miles, and the car's now done 51k miles. How well protected is the connector from dirt/water ingress? My wife drove through what she thought was a puddle but was actually a pool of mud, and the state of the inside of the engine bay is nobody's business (same for the drive when all the mud under the arches dried and fell off!)
3) A source of inexpensive parts! Cheapest I've seen a front Lambda sensor for is 42 on ebay, and 199 for the cat. Many places (GSF included) don't seem to acknowledge the existence of the 1.2 3cyl engines.
4) Is it common for the rear wash wipe pipe to become disconnected inside the loom inside bay next to the gearbox? The squirter didn't work. I stuck my head under the bonnet and got her to wash the rear window, and water started pouring through what appears to be the tape-wrapped loom.
5) As I was removing the engine cover/filter assembly, I realised there was a rubber pipe from the filter box that leads to a cylinder which I *assume* is the EGR (seems to be attached to the right hand side of the exhaust manifold, and then to the inlet manifold by a solid brassy coloured tube, about 2cm in diameter, and looks like the EGR on my car). This tube goes to the bottom end of the EGR, towards the bulkhead.
I don't know the function of this tube though. Could it be letting enough air in, when running in closed loop, to confuse the ECU? I've run the engine (stationary) with the cover off, and there seems to be a slight vacuum when the car's running, but it doesn't run any differently covered or not.
It's also done it with the pipe connected, so I'm assuming it's not the problem.
Many thanks for any assistance.
Pete.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
pete snipped-for-privacy@lethe.org.uk wrote:

snip
my first thought would be plugs, rapidly followed by coil packs
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Tue, 21 Apr 2009 09:57:39 +0100, pete snipped-for-privacy@lethe.org.uk

It shouldn't, you might have an old one but there's a pre cat & post cat lamda on the ones i've seen.
the No2 O2 sensor shows (as far as I can tell from a 3Hz

That's working then.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Motorsforum.com is a website by car enthusiasts for car enthusiasts. It is not affiliated with any of the car or spare part manufacturers or car dealers discussed here. All logos and trade names are the property of their respective owners.