It would be very difficult for the Japanese to actually win the Battle of Khalkhin Gol given both the regional disparity in forces as well as the rules of engagement for both sides. Even if Tokyo opted to allow more reinforcements to be sent to General Komatsubara and lifted the ban on air strikes in Soviet territory, the Red Army and Air Force could simply reciprocate in kind since they had many more reserves on hand. In a contest of attrition in 1939, even a limited one like Khalkhin Gol, the Soviets had more stamina than the Kwantung Army. Probably the only way the IJA could 'win' the battle was if Zhukov pulled back after the Japanese offensive in July (which historically broke into the Soviet rear areas before being stopped by an armored counterattack), which would only have come on direct orders from Moscow.
However:
Gaijin is 100% correct, the effect of Gholkin Kol was to basically make the IJA go "NOPE! FUCK THAT SHIT! NOPE!" when anyone brought up the idea of a jaunt to the North...
...So it don't even help Germany that much apart from go to the Russians "Hey, guys your armed forces are a bit lacking in a few areas..." The Russians wouldn't even need to divert troops from the West, they had an absurd numerical advantage in men, guns and tanks and whilst the T-26 and BT-7 and co would fare badly against Panzers in 41, in 36 they are so far above what the IJA had that the Japanese might as well be fielding AV-7's...
...And as badly trained as the Soviet army was in 36, that are going to slap the IJA round the face with a house brick until it stops being funny if the Japanese fight a battle the way the Soviets want them to.
All of these are 'pop history' memes and need to die. Khalkhin Gol did very little to dissuade Japan from planning to attack the Soviet Union; even into 1944 (yes, while the US was wrecking their carrier fleet in the Marianas and launching B-29 raids on the Home Islands) the IJA envisioned launching a land invasion of Soviet territory should a war have broken out. Additionally, during the actual fighting at Khalkhin Gol both the Japanese tanks and infantry consistently outfought their Soviet opponents: although the BTs were better on paper than the Japanese Ha-Gos, the Yasuoka Group tankers knocked out many more Soviet vehicles in pitched engagements than they themselves lost in return, and each time the Red Army attempted infantry attacks on the Japanese positions they were slaughtered. The worst case of this was the series of probes Zhukov launched on 7/8 August to "feel out" the defenders prior to the big show on the 20th; the combined results of these were over 1,000 abandoned corpses on the Soviet side and several tanks knocked out, whereas Japanese
casualties (not just killed, but casualties) numbered just 85.
On the whole, prior to Zhukov's general offensive on August 20th the battle was largely a stalemate, with the Soviets being on the receiving end of a nearly 3 to 1 casualty ratio (a ratio also present at Lake Khasan, where the Japanese were even more outnumbered and outgunned).
EDIT: It's also worth keeping in mind that the vast majority of sources on Nomonhan are Japanese, and I haven't yet seen a detailed English-language Soviet account of the battle. Even Coox's seminal work relies overwhelming on Japanese sources, despite still painting a grim picture of Japanese capabilities, written as it was a year before the collapse of the USSR. There are known problems with relying on sources from just one side of a battle, particularly with regards to kill claims and the size of opposing forces. If we had the same wealth of tactical sources from the Russian side we might discover that the Japanese did rather less well in battle than even they thought they did.
Coox relies heavily on Soviet-era sources for the narrative on the Red Army side, though most of the book consists of a tactical view of events from the Japanese perspective as described by many of the latter's veterans, either through direct interviews or war journals. The full breadth of information currently available to us from Soviet/Russian sources (Kolomiets, Kondrat'ev, or even
the 2013 publication by the Institute of Oriental Studies as edited by E. V. Boykova), simply was not accessible to him in the 1970s and 1980s. From this limitation the reader gains an impression that the battle was much more one-sided than it actually was.
In reality the Japanese Army's claims of damage inflicted on the Soviet side were significantly
understated compared to the real thing, a situation paralleled by the Finnish Army's claims in the Winter War. According to figures used by the Japanese in the aftermath of the battle, their estimate of Soviet casualties was about 18,000 ("not less" than their own) with 400 AFVs destroyed - the real figures were 27,880 and 386, respectively. The only major overclaim was in the air, where IJAAF aviators reported over 1,200 downed Red planes, more than six times the actual total. The Soviet 1st Army Group, for their part, initially gave Japanese casualties as 29,085, which was much closer to the truth than the 50 or 60 thousand often seen in "official" sources.
Anyway my point about the loss ratios was that IF the Japanese had more forces to commit, which wouldn't have necessarily had to be that much, say an additional prepared division or two and some extra armor/aircraft/artillery, they could have probably won against the historical Soviet forces.
As it was apparently Coox says that the Japanese detected the buildup but did not react.
Though again I'm not that well versed on all the specifics off hand (I haven't argued about this a while) and
@BobTheBarbarian has done ungodly amounts of research on that specific battle and has the numbers.
After the 23rd Division was encircled the Kwantung Army realized what was happening and put together a 'relief force' consisting of three divisions, elements of two more, another tank regiment, 47 37mm AT guns, a motorized mountain artillery regiment of two battalions (24 guns) 34 75mm regimental guns, two 150mm howitzer regiments, three engineer platoons, and 21 transport companies plus some auxiliary railway units with 1,500 vehicles. These finally arrived in-theater on the 8/9th of September, by which point the battle was long over. Even these units, as ObssesedNuker pointed out with reposted IXJac quotes, were still outnumbered by Zhukov's 1st Army Group and did not counter the imbalance in tanks, artillery, and logistical assets. Given the Soviets' losses (another 9,000+ since 20 August) they probably could have thrown them back on the defensive, but by that point though we're just back to paragraph 1 of my post, and the Soviets can sit on their superior supply lines to build up beyond Japan's capacity to respond, counterattack, (rinse and repeat) until the latter either gives up or goes to war. It especially didn't help that the Manchurian rail network was extremely sparse near the Mongolian frontier, forcing the IJA to rely on its limited stock of motor vehicles as it tried and failed to keep up with Zhukov's elite group.
Similarly, if these reinforcements had been there from the beginning it might have allowed the Japanese more tactical success (for example, their offensive in July probably would have succeeded in inflicting a temporary defeat on 1st Army Group), but again we're back to the situation where the Soviets can just rebuild until they have the strength to push back the IJA. As long as the USSR was determined at all costs to hold their claimed border in Mongolia there was little the Japanese could do to attrit them. The only time the Kwantung Army, as a whole, ever possessed a parity or superiority over the Soviet Far East forces was from the summer of 1941 through the first half of 1943, when it was at the height of its power.