This card is the one-time program version of Desperado run economy.

Lets compare this card to Jackpot, a card with similar mechanics. At the start of each turn, you would place 1 credit on Jackpot, but can only retrieve them when you steal an agenda, (which may not even occur late in the game). Bankroll places a credit per successful run, and if we consider that as once per turn for an aggressive criminal, is similar to Jackpot in that regard. However, it also costs 1 credit install compared to Jackpot's 0.

Its credits can be pulled at any time, compared to Jackpot's only when an agenda is stolen, but occupies a memory unit in the meantime. However, Jackpot can be trashed if tagged.

In conclusion, its power level is about equal to Jackpot's, and should be considered in aggressive criminal decks with spare MUs.

Although Tycoon's stats and numbers are above average, they are no where near enough to warrant its negative ability.

Lets consider Demara as a comparison.

Stats: Install, Base Strength, Boost, Subroutines.

Demara, ~0, 1, 2:3, 2:2

Tycoon,,,,,,1, 1, 2:3, 1:2

So, in exchange for giving the corp +2 credits, you save ~1 credit.

In addition, Demara is more powerful due to additional draws being usable for their bypass ability instead of being dead draws, making it even better than Tycoon.

In conclusion, you should just get Demara instead of Tycoon.

I'd like to point out, since you didn't, that this could potentially feed into cards like Diversion of Funds, and I feel it exists for that reason (as well as being extremely efficient). It'll definitely get you into servers in a pinch. —
On the topic, Demara costs 4 credits to install to Tycoon's 1 credit. As Lynx pointed, Tycoon works well with PAD Tap. —
Synergy with Diversion of Funds is rare. In addition, being 1, (sometimes 2) credits more efficient is not enough to warrant its negative ability, (and is definitely not "extremely" efficient compared to Demara). —
However, its synergy with PAD Tap is notable, and may make this usable. —
Also, Demara has a trash ability that allows it to bypass an ice, negating its install cost, (or sometimes even exceeding it), in the eventually net value. —

Lets suppose you are a big rig shaper aiming for R&D lock with click compression, because that is the type of deck you would play Upya in.

Sure Gamble is a popular economic card, in that it provides a profit of 4 credits with a credit requirement most runners have or can comfortably click to. It provides its credits in a burst, providing you with more credits immediately.

Upya is essentially a run economy card, (save for the rare combo turn usage), that is only usable with click compression, (such as Magnum Opus or Professional Contacts).

If you run every 2 turns to compensate for multi-access runs, economic limits, and corp securing of R&D, you will gain 3 power counters in 6 turns. That means you gain 1 click every 2 turns, and with click compression making every click worth ~2 credits, amounts to about ~1 credit every turn, which is... actually an okay power level. In 6 turns it will even with Hedge Fund.

However, for this card to provide that benefit, you would need to have a click compression card and your breaker suite already installed, in addition to its memory requirement. You may also have difficulty in running on R&D.

Another comparison would be to Data Folding, which needs 2 free MU but does occupy them, and evens with Sure Gamble after 7 turns.

What you say is correct for that archetype. But Upya is built for hammering R&D more often than once every two turns! There is Equivocation, Mirror, TTW and so much other stuff around for runners that run R&D much more frequently in the mid game. —
The problem isn't fresh accesses, its being able to afford running on R&D. Once your opponent realizes your goal, they are going to stack R&D deeply, so running on R&D will become very taxing. I am taking into consideration multiaccess tools to circumvent the effectiveness of your opponents counter by minimizing the number of runs required while still maintaining or staying close to R&D lock. —
A deck can only extend its economic strength so far, especially since rotation took off a large portion of those economic options. If the corp player makes R&D 3 ice deep, each taxing 4 credits, then it will cost 12 credits to access card in R&D. A full turn of Magnum Opus will only raise 8 credits, and even if Professional Contacts can raise more, I simply do not see how you can possibly afford to run on R&D (every) turn. —
*cards, (third line). —

In terms of its numbers, Otoroshi has a gross tax of 6, (and a net tax of ~4), which are great in relation to its rez cost of 2. However, Its weakness lies in its subroutine, which the runner may allow to trigger.

Otoroshi can be placed on a remote server to threaten a flatline with Project Junebug or a rig wipe with Neurostasis if it is protecting an advanceable card, (making it ideal in a shell game glacier).

Otoroshi can also place advancement counters on cards that cannot normally be advanced, and can be combined with API-S Keeper Isobel to become an economic booster, making this ice very efficient and impactful.

Other synergies include Trick of Light for fast advance, Back Channels on an advanceable trap for an economic boost, and Constellation Protocol to support advanceable ice and Mass Commercialization.

If you find any other uses or synergies, mention them in the comments.

It synergizes pretty well with False Flag if its protecting it. They will either boost it, allowing you to score it if they cant trash it, or make it more punishing to trash if they were going to do that anyway. —
re: Constellation Protocol "synergy", the advancement counters from Otoroshi can only be placed on cards in servers, and Constellation Protocol can only affect advancement counters on ice (not in servers). It's a nonbo, sadly. —

This card is much more powerful than it seems. You should consider putting this into your shell game Jinteki deck, instead the classic Project Junebug.

The most obvious comparison to make is to Project Junebug, which instead does 2 net damage for every advancement. In game, when playing shell game Jinteki, you mostly install and double advance to maximize the punishment if it is a trap and normally score if it is a 5-point agenda. Project Junebug with 2 advancements would do 4 net damage, trashing most of their hand and slowing down their tempo by making them click back their hand, which would take 1 turn of drawing or 4 clicks. It also has a side benefit of possibly disrupting their game plan.

Neurostasis with 2 advancements would trash 2 installed cards, and if they are icebreakers it would force them to search through their stack and reinstall them, which would secure your centrals and open a scoring window lasting multiple turns while taxing them. Reinstalling with a draw and install, (2 clicks), for both would cost 4 clicks, and that is not factoring their combined install costs, (usually about ~6 credits).

Therefore, Neurostasis is superior to Project Junebug in both taxing and utility.

Neurostasis is super underrated yeah but I think we should refrain from saying it's superior to Project Junebug. Similar to Degree Mill, Neurostasis can actually become a liability when it shuffles back drained Davids or Liberated Accounts. Don't forget the corp is forced to choose runner cards (assuming they pay) and so Neurostasis is very nearly a meager card in the early game in comparison to Junebug or a similar Ambush. Not to mention Neurostasis' 3 to fire compared to Junebug's 1. —
The runner would not be running on a scoring server without icebreakers, especially with how punishing Jinteki ice tends to be, —
The second negative would only apply if you are naked advancing it, which is what Saraswati does. —
As for the first negative, being forced to uninstall, usually does not happen. —
Paying 2 additional credits for a long lasting scoring window is very valuable. —
However, you are right that Neurostasis is not nearly as useful in the early game. —