It was well known that Tetraethyl Lead was toxic by the time that it became an additive to fuel for knock resistance.
After some deaths in a production facility, sales of TEL were suspended for a year and a conference held into its health issues. That, and other studies that were held in the following years, sometimes conducted by the lead industry, found that there was little health risk to the public.
But what if warnings from a few scientists were heeded and TEL sales permanently banned from 1925?
The other anti-knock additive being proposed at the time was ethanol. Had that been adopted what would aircraft performance have suffered?
How would ranges have been affected by petrol/ethanol blends? Fuel economy is ~3-4% worse with E10 compared with regular petrol.
It's worth noting that the acute toxicity of TEL was its best understood aspect, as was acute lead poisoning (and acute heavy metal poisoning in general, as with mercury). It's the chronic, low level lead exposure that was more poorly understood and being investigated and contended at the time. (particularly the fine particles of lead in exhaust from leaded gasoline)
That said, had the full dangers been understood, I suspect at most this would've led to restrictions on automotive gasoline, eliminating the vast majority of public exposure (and contamination of soil and ground water) presented by leaded fuel.
By the 1920s, a variety of octane boosting (and high octane fuel components) were available and it was mostly down to automobile manufacturers to design the engines and fuel systems to tolerate a wider range of blends with alcohols, aromatics, and other fuel additives or alternative fuels. It shouldn't have been too hard to find some compromise and set standards on the fuel/air mixture limits, lubricity, water tolerance, solvent properties (erosion or swelling issues for rubber based seals and tubing), volatility, flash point, etc for general automotive use and additional standards for different seasons and different regional climates. (regions with higher humidity might have stricter limits on water-attracting methanol and ethanol and/or higher requirements of cosolvents to prevent phase separation)
Also note: catalytic converters are a non issue here and restrictions of sulfur emissions (and acid rain) would likely be of limited concern, so the sulfur containing fraction of crude oil and resulting gasoline that has both octane boosting qualities and good lubricating qualities could be explicitly retained to improve those characteristics.
Between the development of new synthetic processes to produce methanol by the late 1920s and an increasing array of liquid alcohols, ethers, ketones, and hydrocarbons using synthesis gas as a starting point, plus the ongoing advances in catalytic cracking of crude oil into higher yields of desirable high octane components, and the competing fields of biofuels (namely fermentation derived ethanol and acetone-butanol-ethanol fermentation) all should've helped work towards a quite viable set of standards to work with during the 1930s and into the 1940s.
Also note: certain combinations of the above components can also mitigate issues with metallic corrosion or erosion of rubber (plus corrosion inhibators can be added as well) plus combinations of light alcohols like ethanol, and especially methanol with normal gasoline (and hydrocarbons in general) will actually increase volatility and lower the flash point beyond that of either of the starting materials due to the semi-polar nature of the alcohols vs non-polar hydrocarbons, plus the formation of azeotropes with several common hydrocarbons present in gasoline (like hexane, plus aromatics and benzene), increasing the vapor pressure and lowering the boiling point of the mixture. At some point adding more alcohol will stop increasing the volatility and start reducing it (usually somewhere between the 40 and 60% alcohol mark) and taper off down until you reach the same vapor pressure of 100% alcohol. (note: methanol is more volatile in itself and more aggressive in increasing volatility of hydrocarbon blends both due to its lower boiling point and more polar nature than ethanol, albeit this also means it can phase separate with lower water content than ethanol and thus benefits more from having ethanol, propanol, and/or butanol present, or other co-solvents, to reduce this behavior)
This behavior is also partly why it's uncommon to see E20 to E50 blends in modern gasoline since the EPA has fairly strict limits on vapor pressure and those intermediate blends exceed this. (this was also a major point holding back methanol blends tried in the 1970s and 80s in the US, including some that were initially granted waivers and later revoked, namely concerns over VOCs emitted by fuel evaporation from the tank). This is a bit frustrating and ironic as it's these blends that are most efficient and tend to greately improve break thermal efficiency (to the extent that many engines have better per-gallon fuel economy than with straight gasoline, with 30 to 50% generally being optimal, though much over 35% not being tolerated by typical non-flex-fuel engines simply due to inability to correct the fuel/air mixture and ending up running lean, thus triggering a fault from the oxygen sensor). Much beyond 50% and the efficiency drops again, and by the time you get to E85 the brake thermal efficiency is much closer to the same a sit is with normal gasoline or E10.
Alcohol blends also reduce flame temperature and allow engines to run cooler.
Ethanol and methanol are also especially corrosive to aluminum (and many aluminum alloys), but interestingly this is most true for anhydrous (water-free) alcohols which can dissolve aluminum oxide and form a gel of aluminum hydroxide and alcoholate that accelerates further corrosion so long as any dissolved oxygen is present. However, a small fraction of water is enough to inhibit this, with less than 1% being needed to achieve this and prevent dissolving of the protective oxide layer of aluminum. Of course, the presence of water itself can have other corrosion issues (and phase separation issues), moreso for steel components than aluminum, though there may be a happy medium for very low, non-zero fractions of water, possibly even within the convenient hydrous (190 proof, ~95%) ethanol range.
Experimentation with different steel and aluminum alloys, various fractions of water, different fuel blends, and corrosion inhibitors likely would've worked around this, along with maintenance procedures specific to these issues, especially for extended periods of inactivity. (problems from fuel sitting in float bowls and such causing issues)
On the plus side, alcohol blends and acetone tend to greatly reduce gum and varnish problems in fuel (where aromatics tend to make this worse), and the intermediate alcohol blend proportions that tend to optimize fuel economy also tend to have some of the best water tolerance. (small amounts of water would also further increase octane rating without, plus further reduce flame temperature, allowing engines to run even cooler than the alcohol blend already allows)
Such blend may have been appealing enough to even use for aircraft during war-time given the higher octane and potentially competitive fuel economy, though real-world experimentation would've been needed to confirm this and the higher maintenance requirements (and potential inability to allow unsealed fuel systems to sit with fuel in them) would've been concerns for anything involved in front line combat service. However, the lubricating qualities of TEL would still likely be desired for aircraft use along with the octane boosting quality, albeit TEL tends to be less effective at boosting octane of alcohols than hydrocarbons, this also varies somewhat with blending. OTOH a minimum amount of TEL required just for the valve lubrication and wear qualities should've allowed stretching of limited TEL supply during war-time as well.
The earlier you put pressure on designing engines and fuel systems around alternative/competitive fuels, the fewer problems with logistics you'll have later on. And, given benzol and ethanol were the earliest competitive alternatives to TEL for automobiles (and to high octane straight gasoline from California oil wells and the like), optimizing around a range of alcohol-benzol-gasoline blends would've made sense, and from there you'd already have a lot of the general difficulties being worked on, so a wider array of blends later on wouldn't change all that much. (ie the difficulties with just benzol and ethanol would account for the vast majority that would be encountered with other alternative blends that might come later as synthetic chemical technology expanded) Acetone and n-butanole available through fermentation following WWI would also have been early considerations (the ABE process developed in the UK as a source of acetone for cordite created a new bio-chemical industry for ABE fermentation in the US that lasted up through the 1950s when access to cheap sugar cane from Cuba was cut off).
And, of course, during prohibition in the US, you have a further surplus of industrial alcohol (and underutilization of distilleries) that would've made ethanol fuel blends more attractive to all segments of the agriculture and alcohol production industries otherwise marginalized by the lack of (legal) demand for ethyl alcohol in consumables.
Racing aircraft throughout the 1920s and 30s had been experimenting with various alcohol fuel blends, ethanol and methanol both, plus benzol and acetone, and the Italians used ethanol blends for at least some of their fighter aircraft during the 1930s. The Fiat CR.32 and its A.30 engine, apparently ran on a mixture of: 55% petrol, 23% ethanol, and 22% benzol.
Additionally, aircraft using pressure injection carbs (single point fuel injection) or direct injection would've avoided the issues associated with fuel sitting in carb float bowls, exposed to the atomosphere. (and sealed fuel systems of that sort should've been easier to allow for alcohol fuel blends to sit in the tanks and in the fuel lines for extended periods between operation)