There are different size filters to accomodate velocity stacks. It sort of depends upon the stacks themselves, how tall they are and their diameter.
Most Kastner stacks I've seen are "full radius" design and would require a pretty large filter box, of the usual designs, possibly more than the engine compartment can accomodate. Sock style filters, one per stack, might be the only choice. I agree with BobH that this probably isn't the best method of filtering, if there are alternatives.
With K&N and others similar filters that have a solid cover, there should be a min. of an inch clearance, between the outer edge of the velocity stack and the inside face of the filter cover, a little more would be even better. I've got 3 or 3.35" deep K&N filters fitted over 40mm stacks on 40DCOE Webers. That should be adequate clearance.
There might actually be some advantage to running velocity stacks under filters. To be as effective as possible, air around the carb intake/stack should be relatively still. I believe filters will help with that. This is why many carbs fitted with velocity stacks often have a roomy air box covering them.
Those you see pictured without any sort of air filtration are either full race engines that get rebuilt every couple thousand miles, or are for show. I wouldn't run a street car all the time without some filtration to protect the engine. There is just too much danger of damage to the engine, as BobH described.
There are two primary purposes to velocity stacks, besides looking and sounding cool. One is to "clean up" air flow and channel it into the carb. The other is to capture "stand off" fuel/air mixture that might gather around the mouth of the carb, that might be blown away without the use of stacks.
If those velocity stacks truly increase flow in your engine, you may need to enrich the carbs slightly to dial them in. In other words, adding the stacks to a carb tuned to work without them should cause mixture to be a little lean. Best test might be an exhaust gas analyzer up the tailpipe of a fully warmed up engine, tested at various rpms, after the stacks are installed.
Overall, velocity stacks seem to be useful on DCOE and certain vertical hihg performance Webers. There are various lengths available. With Webers, a longer stack has more effect on low rpm range, shorter is more effective at high rpms.
I've heard, but not seen any real proof, that SU efficiency sees little improvement with velocity stacks. In fact, I've seen advice to use only very short or to simply shape a radius on the inlet side of the carb itself, to get about as much benefit as possible.
ZS carbs are very similar in design and function to SU carbs, so maybe this info applies to them as well. Then again, maybe not.
Velocity stack effectiveness also depends on other tuning that's been done to the engine, increasing the engine's breathing in terms of volume and the speed of the fuel/air mixture. An engine with porting, gas flowing, increased compression, headers, a hot cam and other improvements will see more benefit from adding a set of stacks of the right length. How much? Well, likely just a little.
Racers do a lot of tweaks and modifications trying to get the last little bit of performance out of an engine, with no regard for long term reliability and street manners. Unfortunately, a lot of cool "racy" stuff ends up installed on street cars where some of it has little or no benefit, might actually cause problems in the long run. I don't think velocity stacks on a street car are likely to do any harm, so long as a filter is still used. But, I also wouldn't expect a big performance increase.
Heck, have fun, give em a try and let us know what you think!