I'm not persuaded that the instances of imflammable over the years are the result of a conscious effort to establish a less confusing variant of inflammable—as opposed to being a longstanding variant with no pretensions to reducing the number if incidents of accidental flash fires.
'Imflammable' as a less easily misunderstood variant of 'inflammable'
Truth to tell, I couldn't find much direct information on this subject, except a comment that appeared in various publications of the American Society for Testing and Materials Solutions between 1960 and 1983 urging people to stop using imflammable and other variant spellings—presumably because these variants were only filling the room with smoke. From American Society for Testing Materials, ASTM Standards on Textile Materials: (with Related Information) (1960):
NOTE.—It is recommended that the use of the terms imflammable, inflammable, nonflammable, nonimflammable, and noninflammable be discontinued.
The following terms are used in the references indicated :
Incombustible—Proceedings, Am. Soc. Test. Mats., Vol. 45, p. 866 (1945).
Fire Retardant—Tentative Specifications for Fire-Retardant Properties of Treated Textile Fabrics (ASTM Designation: D626).
Flameproof (Fire Resistant)—Federal Specification CCC-D-746. 1956 Yearbook, Am. Assoc. Textile Chemists and Colorists.
In short, the only member of the inflammable family of words to survive the comprehensibility conflagration resulting from efforts to avoid confusing people as to the meaning of inflammable was the word flammable.
An earlier note to similar effect appears in George Haven, Industrial Fabrics: A Handbook for Engineers, Purchasing Agents and Salesmen (1949) [combined snippets], which itself drew from the A.S.T.M. Standards of 1947:
Flammable, adj. — Highly combustible, producing flash fires. Flammable materials, when ignited, ten to produce flash fires because of their physical state or condition. The latter is generally associated with a relatively large surface area or a high vapor pressure. Specifically a flammable material is one which will support a flash fire at a rate greater than that specified in a standard test.
Imflammable and Inflammable, adj. — Flammable. It is recommended that the use of these terms be discontinued because of possible confusion in the meaning.
Nonimflammable, adj. — Nonflammable. It is recommended that the use of this term be discontinued.
At this stage, it seems that the ASTM was still plumping for nonflammable to be adopted as the standard way of saying "not inflammable"; but by 1960 the organization appears to have thrown the sponge into the bucket on that idea.
But if imflammable was already under a dark cloud—and indeed being recommended for excommunication—in 1947, when and by whom was it being promoted as a way forward for people trying to avoid the ambiguity of inflammable?
Multiple 19th-century instances of 'imflammable' in the same article
There are a couple of instances from the nineteenth century in which imflammable appears twice on the same page, lessening the possibility that its occurrence was a simple typo, but perhaps increasing the possibility that it was written by someone who didn't know how to spell inflammable. From a letter to the editor of The Belfast Monthly Magazine (September 1809):
By a number of chemical experiments it has been ascertained that the basis of all earthly productions is the carbonic acid, or fixed air, and heat or fire ; these compounded with hydrogen or imflammable air, produces the essential oil, with which vegetables abound, consequently soils which contain the essential oil, are the best adapted for cultivation.
Hydrogen has its name from its property of producing water, it is also called imflammable air, from its property of burning or exploding with vital air ; it is not proper for respiration, yet it has not the sudden mortal effects of nitrogen.
From "The Texas & Pacific Railway Co. v. Levi & Bro.," in The Texas Law Reporter (1883):
If the same lot was used as a lumber yard, in which to store plank and other imflammable material, it would be a question of doubt, as to whether such use, reference being had to the use to which contiguous property was lawfully appropriated, was negligence or not. So in the use of a yard, contiguous to a railway track, for the purpose of storing baled cotton, which as matter of common knowledge is imflammable and easily ignited, when it is shown that contiguous property is lawfully used for a purpose from which it necessarily results that the cotton will be subjected to some danger from fire, it might be doubtful if the storing of baled cotton in such a place was a prudent act, and the question should be submitted to the jury.
From "Is Fine Wood Dust Explosive?" in The Wood-Worker (August 1887):
There must, however, be some occult and peculiar conditions present in order to make wood dust explosive, because we have often seen, in a cedar bucket factory, what we consider the most favorable conditions for an explosion of this kind. Dry red cedar is the most easily imflammable of all woods, and in every large factory the air in the finishing room is loaded with impalpable cedar dust, and every time the sandpaper is applied an intense heat is generated, the blue, hot smoke and the finest and most imflammable of wood dust arise in one mixed volume.
And from a rule change enacted on September 2, 1896, in Votes and Proceedings of the Legislative Assembly [of New South Wales] During the Session of 1896 (1896):
Rule 7. If at any time it is found by the person for the time being in charge of the mine o any part thereof, that by reason of imflammable gases prevailing in the mine, or that art thereof, or of any cause whatever, the mine or that part is dangerous, every workman shall be withdrawn from the mine or part so found dangerous, and a competent person appointed for the purpose shall inspect the mine or part so found dangerous, and if the danger arises from imflammable gas, shall inspect the mine or part with a locked safety-lamp; ...
Presumably the 1809 instance of imflammable did not represent an effort to avoid having readers misunderstand the meaning of inflammable—but it is hard to tell whether that sort of thinking might have had some influence on the three instances from 1883–1896, a period when the spelling simplification movement and other efforts at orthographic rationality were increasingly popular. I have not, however, been able to find a manifesto on behalf of imflammable as a spelling less likely than inflammable to lead to tragic misinterpretation.
Twentieth-century publications that show strong signs of having intentionally used imflammable (because it appears more than once in the article) include Telephone Cables (1906), Engineering and Mining Journal (February 18, 1911), American Municipalities (November 1916), Mining Magazine (1921—using the word non-imflammable twice), and Ice and Refrigeration (July 1922). The only multiple instances of imflammable that I've been able to locate in a book published after 1922, however, are from Scientific Research Abstracts in Republic of China (1989), a source that may very well have used that spelling twice in the same paragraph without realizing that it was odd.
Just because I didn't find a contemporaneous argument touting imflammable as a safer spelling than inflammable doesn't mean that no one was making that argument—1n the 1880s, the 1910s, the 1920s, or later. But until I see such an article from that era, I remain skeptical that any such argument underlies the more frequent usage of imflammable during those decades.