I cannot claim much hands on expertise in this area - I have only ported/chambered two heads according to Vizard's suggestions and no flow bench or dyno to measure actual changes made. It seems there are a number of quite (very?) complex relationships between flow potential (larger ports), gas speed (smaller ports), distribution (depends to some extent on carb setup), scavenging effects (depends on cam), fuel atomisation and surface 'wetting', and a couple of elements unique to the 5 port 4 cylinder engine. TK is quite right - just getting stuck in with a grinder is unlikely to make improvements across the operating range of the engine, although may help some part of it. If you look at GR's heads, you can see the amount of thinking that goes into a number of very subtle mods that 'add up' to a significant advantage. And even the experts occasionally screw it up - only takes a pinhole leak where the casting on that particular head is thinner than usual. However, doing a bit of smoothing through the ports, some careful attention to de-shrouding the valves (particularly on the 'closed' head designs), and the valve throat/pocket, and a gentle rounding on the dividing 'beak' between the inlet valve throats (to allow smoother transference of flow) should give you some advantages. The exhaust ports are often smaller than the gasket openings or the extractor openings. This needn't be a big problem, as smaller ports keep gas speed up and the resulting lip can have an anti reversion effect for cams with a lot of overlap between the intake and exhaust phases. My suggestion, for what it's worth (which is much less than the real experts) is that if you want a really good high flow head, buy one from someone who knows what they're doing. If you enjoy a bit of tinkering and adding your own contribution to the mix, some gentle smoothing of the ports will aid flow - but at your own risk. If you can, find a clapped out head to practice on before you take the grinder to your good one.
