My vote: None. Most of them are useless, all of them are a serious money drain and a hassle to store and use, and those few that do kinda sorta almost work create other problems (combustion chamber deposits that aggravate pinging, spark plug fouling). The one or two products that contain actual tetraethyl lead is
still on the market, but you have to be seriously messed up in the head to futz with this stuff. It is
hideously, grievously toxic to handle, store, and be around. It is also ridiculously expensive. Any reasonably tame street engine driven in normal service will do fine for quite awhile on unleaded fuel; this is especially true of slant-6s and 318s with conservative (i.e., small) valve sizes. Run the engine til it needs a valve job, then have hard seats put in and the problem completely ceases to exist even as a theoretical concern.
There's a fair amount of misunderstanding and myth on the matter of lead in gasoline. Here, allow me to babble at some length about it:
The mechanism by which lead prevented valve seat recession is commonly misunderstood. It's nothing to do with lubricating anything. With an unhardened exhaust valve and seat, the valve and seat can micro-weld to each other if they get hot enough. Lead acts as a buffer to prevent this happening. The important thing is that exhaust valve and seat recession only takes place when the valve gets hot enough to undergo localised welding. Then, when the valve opens next, the metal pulls apart like taffy. This roughens the meeting surfaces, and they become quite abrasive. The pounding/turning of a valve with such "pulled" metal on it creates a nice grinding wheel effect on the seat. In addition, the roughened surfaces no longer seal against each other properly, which eventually allows still-burning combustion gases to flow through the "closed" valve, causing a blowtorch effect on the poor valve and depriving it of any prayer of a chance to cool while it's on the seat. The blowtorch effect rapidly deteriorates the seal further, snowballing the seat recession.
The main thing to remember is that this bad stuff cannot happen if the valve never reaches the crucial temperature. Whether the valve reaches the crucial temperature depends mainly on how the car is driven and used (Towing, drag racing or pedal-on-the-floor hauling will heat the valves; driving down the highway at a constant 70 won't, and neither will hopping from traffic light to traffic light in the city or running down to the local grocery for a carton of ice cream.) Other factors in the margin of safety include the size of the exhaust valve, its material, and the efficiency of valve seat cooling in that particular engine design.
Very little lead is required to prevent the localised welding and "taffy pull apart" effect that leads to the abrasive surface which, through incidental or positive rotation of the valve, eventually grinds down an unhardened seat. The majority of the lead was in yesterday's fuel because it was a very cheap and very effective octane booster. When unleaded fuels were first widely introduced (by legislation), there was generally only one grade of unleaded available, and the octane was quite low, less than that of leaded regular. We all know that when you use a fuel of insufficient octane, your engine pings. This creates tremendous heat in the combustion chamber, certainly enough heat to push the exhaust valves to the crucial temperature. Because for quite a while only unleaded fuel of sub-regular octane was available, plenty of people experienced these effects from using unleaded. While many of those engines that suffered under this low-octane unleaded really did need the lead (high load and/or high-RPM engines), the bulk of the failures were due to the low octane increasing combustion chamber temperatures. And so the myth was born that old cars' engines will quickly die if run on unleaded.
These days we have wide availability of high-octane unleaded fuels, which obviate the insufficient-octane cause of valve heating and subsequent localised welding. if you have an old car that is a low-stress application , used in daily-driver service, then you need have no qualms about using whichever octane grade of lead-free fuel your car runs well on and drive it for a long time with nary a valve or seat problem. Many US 6 and 8 cylinder engines fall into this category in normal daily driving service.
The way to eliminate even the possibility of valve heating causing localized welding and subsequent seat recession is to install hardened exhaust valve seat inserts and higher-grade exhaust valves. This is utterly standard practice in the rebuilding of cylinder heads, and has been for years. Hard seats and valves are readily available for just about anything you want to put them in. It's a very common operation and a competent machine shop can handle it. But the main thing here is that there's absolutely no reason to tear into the engine solely to install hard seats. There is no collateral damage from seat recession. Drive and enjoy! You likely won't experience any problems for a long time.
The bottled additives available on shelves vary widely in what they do. Some of them use a sodium salt and claim to duplicate the buffer effect of lead. Some of them use "MMT", methylcyclopentadienyl manganese tricarbonyl, which is toxic, is known to crud-up combustion chambers (more pinging), and is of highly questionable benefit in buffering exhaust valves. Regardless of whether any human-health or environmental risk is posed by MMT, the stuff causes hard red deposits on your spark plugs that will cause you to need to replace them more often (yet more money). Other additives are simply octane boosters of varying effectiveness and varying side effects.
Be careful if you go shopping for these; when they talk about raising the octane of the fuel by so-and-so many "points", they are talking about the number to the
right of the decimal point. That Lead Supreme 130 stuff, for example, says 18 ounces added to a gallon of fuel will raise the octane by "11 points". That sounds like it means you can turn 87-octane fuel into 98-octane fuel, but it really means you're turning 87-octane fuel into 88.8-octane fuel. And you're spending $8.44
per gallon to do so. Does that sound cost-effective to you
?
It's worth noting that on the East Coast of the US, Amoco marketed unleaded high-octane gasoline for decades before the EPA decided to "unlead" the country's fuel by regulation. That Amoco unleaded was widely regarded as quite a fine fuel indeed, and many motorists used it on a regular basis with no ill result.